AI in Design 2026 report v1

# AI in Design Report 2026 Source: AI in Design 2026 --- # Home Meta description: The second annual report by Designer Fund and Foundation Capital on how design teams are adapting to AI across tooling, craft, and org. How to cite this report: Source: AI in Design 2026 | AiiD 26 | Read the Report + | About | Case Studies + | | --- | --- | --- | --- | How designers are evolving their tools, craft, and teams with AI # AI in Design Report 2026 BY DESIGNER FUND IN PARTNERSHIP WITH FOUNDATION CAPITAL Scroll to read OUR PARTNERS AN INFLECTION POINT In 2025, designers were experimenting with AI. In 2026, they're rebuilding around it. - **900+** — Designers surveyed in 60+ countries - **25+** — Interviews with practitioners and leaders AI in Design 2026 aims to capture how AI is transforming tech design across designers' desks and within their teams. We ran our first AI in Design survey in early 2025 because we consistently heard designers and leaders ask, "How are others doing this, and what's working?" A year later, we're attempting to get a sense for what's changed and share firsthand perspectives. The answers come from over 900 designers at startups, enterprises, and agencies who work across disciplines like product design, brand design, research, and design engineering. We also conducted over 20 interviews with leaders at companies actively navigating this shift. Given how quickly practices are evolving, we'll continue to release new findings throughout the year, including case studies about design at companies like Anthropic, Sierra, Stripe, Notion, Shopify, Linear, and Framer. Sign up for new releases. ### 01 TOOLS The great toolstack shakeup AI usage has surged, but the toolstack is still in flux. Designers are using double the number of off-the-shelf tools than they did in 2025, and they're building custom software with AI that matches how they like to work. As everyone rushes to keep up with new releases, reliable output quality remains the largest area for improvement. In this chapter, we'll cover: The most-used AI design tools How the average toolstack has more than doubled What makes tools stick (and why many still don't) Designers as builders of their own tools Tool fatigue and the pressure to always be learning ### 02 CRAFT Craft in the age of infinite output Everyone is shipping faster. But is speed good for craft? AI has unlocked a new gear for designers: they're ideating faster, prototyping more, and learning to code. Half of respondents have pushed AI-generated code to production. At the same time, we hear concerns about craft atrophy and the loneliness of designing alongside AI instead of teammates. In this chapter, we'll cover: Coding as a core design skill Prototyping as a default output The tension between speed and quality Preserving judgment, taste, and skill development The confidence that comes from being a builder ### 03 TEAMS Redesigning the design org Companies have stepped up their support for AI adoption, but most of the learning is still happening between peers. The organizations seeing the most momentum are creating the conditions for tinkering. They're also rethinking collaboration rituals for a world where anyone can spin up a prototype, but the AI tools they're using haven't yet been designed for multiplayer work. In this chapter, we'll cover: How companies support AI adoption Role blur between design, PM, and engineering The messy nature of collaboration Changing expectations and company policy What hiring managers are now looking for Video case studies # Seven companies. Seven ways of navigating the same shift. ### Stripe Creating the conditions for adoption How do you build a culture that enables AI adoption instead of dictating the playbook? Coming soon → ### Sierra Scaling craft, keeping the bar How does a small design team supporting 100+ engineers protect craft without lowering the bar? Coming soon → ### Anthropic When code is no longer the constraint When AI writes most of the code, how should the design team operate? ### Shopify Growing as a designer In a moment when the tools, skills, and expectations are all shifting at once, how can designers grow and thrive? Coming soon → ### Notion Working alongside agents What does it feel like to design alongside agents when the tools and workflows change but the philosophy doesn't? Coming soon → ### Linear Protecting the space to explore As AI accelerates everything, how do you defend the slower, more deliberate parts of the design process? Coming soon → ### Framer Designing the design tools When most AI tools optimize for generation, what does it look like to build one that protects design control instead? Coming soon → ### Inside AI-native design teams Seven video case studies with the design teams at Anthropic, Framer, Linear, Notion, Shopify, Sierra, and Stripe. Go inside the workflows they've rebuilt, the tradeoffs they're navigating, and how they're operating differently as a team. # Get new case studies & report markdown Download the markdown version of the report, ready to drop into any tool. Get notified as new case studies go live. [Email ] [Submit] # About Meta description: The second annual AI in Design report, built on 900+ survey responses and 25+ qualitative interviews with design leaders. | AiiD 26 | Report Chapters + | Case studies + | About | | --- | --- | --- | --- | ABOUT THIS REPORT AI in Design 2026 is the second annual research project from Designer Fund and Foundation Capital on how design teams are working with AI. Last year, we launched the [first edition](https://www.figma.com/deck/F6P0qVl52W3ngNqyNQCGrl/State-of-AI-in-Design-Report-2025?node-id=1-3307&t=YTGuPRSWTCYKqkOx-1&scaling=min-zoom&content-scaling=fixed&page-id=0%3A1) of the report—one of the most widely shared resources in 2025 about AI's impact on design in tech. Now, after the releases of new models and a swath of new tools, we ask the same questions: Which AI tools are design teams using most? What does a typical design workflow look like today? Are hiring criteria and org structures shifting in any meaningful way? We wanted to find concrete, usable answers. This report was created for design leaders making decisions on tooling, policies, and adoption; designers thinking about how their craft and careers should evolve; founders and product teams looking to understand how AI is changing how design gets done; and students and educators preparing for what's next. RESEARCH SUPPORTED BY This research was made possible by our partners, seven companies that shared how their design teams are working with AI today. Each partnered with us on an in-depth case study, created through filmed interviews and live workflow demos. [alphabetical order] Anthropic Framer Linear Notion Shopify Sierra Stripe Anthropic AI LAB AI safety and research company working to build reliable, interpretable, and steerable AI systems. Framer SITE BUILDER The site builder designers love and teams rely on. Linear PRODUCT DEVELOPMENT The product development system for teams and agents. Notion WORKSPACE The AI workspace that works while you sleep. Shopify COMMERCE A commerce platform that helps businesses sell online and in person. Sierra CUSTOMER EXPERIENCE Sierra helps businesses build better, more human customer experiences with AI. Stripe PAYMENTS Financial infrastructure that powers payments for businesses across the internet. Founded in 2013 Designer Fund invests in early-stage startups that use design to improve health, sustainability, and prosperity for all. We back founders who treat design as a strategic advantage and help them build world-class teams, craft standout products and brands, and connect with a powerful community of designers. Our investments include Stripe, Notion, Gusto, Linear, Commure, and Framer. Founded in 1995 Foundation Capital is a first-check investor built to support extraordinary technical founders as early as possible—often before there's revenue, product, or a line of code. Driven by deep conviction, we partner from day zero, helping founders make defining early choices and navigate their toughest challenges. For thirty years, we've stood beside builders who envision futures that seem impossible until they become inevitable. MADE BY This project was created through collaboration between researchers, writers, and designers across multiple teams. | Project lead | Robyn Park | | --- | --- | | Executive producers | Ben Blumenrose, Steve Vassallo | | Editorial | Nathalie Arbel | | Research and data analysis | Nili Metuki | | Identity and website design | ++hellohello | | Creative direction | Heather Phillips, ++hellohello | | Video production | Seed Stories | | Program support | Jackie Berardo | | Media and distribution | Kirsten Underwood, Will McTighe | METHODOLOGY Survey 906 designers in 60+ countries responded to our 2026 survey, fielded in March 2026. Respondents represented a mix of company stages, team sizes, disciplines, and seniority levels. 60% have over 10 years of professional design experience. The survey was created, distributed, and analyzed independent of our partners. We shared it via our social media channels, our email newsletter, word of mouth, and broader networks of online communities for designers. Because of our distribution strategy, the sample may skew toward respondents who have embraced AI more than the average designer or who work at companies that are actively supporting adoption. We present these findings as directional rather than as absolute benchmarks. **Chart title: Survey respondent distribution by work environment** *Question type: Single-select; percentages sum to ~100% because respondents could select only one option. n=906.* | Role level | Percentage of respondents | | --- | --- | | In-house | 71% | | Freelance | 16% | | Agency | 12% | **Chart title: Survey respondent distribution by role level** *Question type: Single-select; percentages sum to ~100% because respondents could select only one option. n=906.* | Role level | Percentage of respondents | | --- | --- | | Individual contributors | 53% | | Managers/leads | 19% | | Executives | 14% | | Founders | 12% | **Chart title: Survey respondent distribution by type of design work** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply. n=906.* | Type of design work | Percentage of respondents | | --- | --- | | Product design | 86% | | Brand and marketing design | 37% | | UI/UX research | 32% | | Design engineering | 23% | **Chart title: Survey respondent distribution by company size** *Question type: Single-select; percentages sum to ~100% because respondents could select only one option. n=906.* | Number of employees | Percentage of respondents | | --- | --- | | 1-50 employees | 42% | | 51-500 employees | 25% | | 501-2,000 employees | 12% | | 2,000+ | 22% | Interviews 25+ in-depth conversations with design practitioners and leaders Conducted February-April 2026. We selected interviewees at companies with a range of stages, team sizes, and approaches to AI adoption. About one-fifth of the interviewees work at companies who have supported this report as official partners. The report also features anonymous quotes from our survey, and are identified only by the respondent's role level and company stage. Companies interviewed: Anthropic, Stripe, Linear, Notion, DoorDash, Cursor, Samsara, Watershed, Airtable, Ramp, Grammarly, AirOps, Abridge, Flux, Framer, Maze, Miro, Shopify, Sierra, Substack, Superhuman External sources 50+ public sources reviewed for patterns and context We reviewed articles, posts, and newsletters from designers, engineers, founders, and hiring leaders to triangulate findings and surface emerging themes. For press inquiries, partnership opportunities, or questions about the research: hello@designerfund.com # Tools 01 Tools ## The great toolstack shakeup While AI usage has surged, most designers still don't have a set of go-to tools A year ago, designers in tech were largely experimenting with AI tools. These showed up in ideation and prototyping but rarely made it to production workflows. In 2026, AI is used across every phase of design work, with 91% using AI for design tasks at least weekly, up from 54% in 2025. The average toolstack has more than doubled, from 3 tools to 7. And the tools themselves have turned over: Claude has overtaken ChatGPT, Figma's role has shifted, and coding tools have rapidly emerged as a core part of the design process. But while the individual tools are becoming more capable, the stack itself is becoming less stable. Nearly half of designers say they're still searching for their go-to tools. And it's easier to build custom software, making the stack much more fluid than we've ever seen it. This raises the question: Will we ever go back to a "standard" design toolset? This chapter maps what designers are using, how AI usage varies by company type, and what makes tools stick (or not). Tool usage The 2026 stack What makes tools stick Designers as toolmakers Confidence vs. noise Where do we go from here? Key takeaways Further reading - **91%** — use AI weekly, up from 54% in 2025 - **7** — average tools per designer, up from 3 in 2025 - **78%** — use Claude, overtaking ChatGPT at 65% **TOOL USAGE** # 1. Frequent AI usage jumped from 54% to 91% in one year 91% of respondents now use AI in their design work at least weekly, up from 54% in 2025—a 37-point jump year over year. 75% of surveyed designers use it daily. **How often designers use AI in their design work** *Question type: Single-select; percentages sum to ~100% because respondents could select only one option.* | Daily | 75% | | --- | --- | | Several times a week | 16% | | Once a week | 4% | | A few times a month | 3% | | Rarely or almost never | 2% | **Regular AI usage among designers (at least weekly)** *Question type: Question type: Single-select (yes/no); percentages show the share within each group who answered yes. Compares 2026 to 2025. Year-over-year comparison.* | 2026 | 91% | | --- | --- | | 2025 | 54% | ### Designers are using AI end-to-end AI has moved from "we use it somewhere in the process" to "we use it at every step." In 2025, the designers we surveyed were mostly using AI in the exploration and creation phases. Only 39% used it in delivery. But today, every workflow we asked about shows meaningful AI adoption. The leading use cases are ideation, prototyping, and UI copy. And in our interviews, designers described continuous AI use across a project, rather than confining it to any one stage. The biggest year-over-year movers include code generation, documentation, design QA, and developer handoff. These workflows sat closer to the edges of design practice a year ago. Most notably, designers in our interviews often mentioned their newfound ability to ship code and build high-fidelity prototypes. 50% of survey respondents say they've shipped code to production (read more about AI coding in Craft). **How designers use AI in 2026, and what's changed since 2025** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply. YoY shift compares to 2025; blank cells indicate workflows not measured in 2025.* | Task | Percent of respondents | Year-over-year shift | | --- | --- | --- | | Ideation and concept exploration | 76% | | | Prototyping | 74% | | | UI copy and content writing | 70% | +18 pts | | User research and synthesis | 56% | +9 pts | | Code generation and front-end implementation | 50% | +31 pts | | Documentation and specs | 50% | | | Image or video generation | 44% | +9 pts | | Wireframing | 43% | + 27 pts | | Stakeholder presentations | 33% | -4 pts | | Design systems and components | 33% | +24 pts | | High-fidelity visual design | 29% | | | Design QA and accessibility | 27% | +16 pts | | Developer handoff | 23% | +12 pts | **THE 2026 STACK** # 2. The designer's AI toolstack has more than doubled Designers we surveyed now use an average of 7 off-the-shelf AI tools, up from 3 last year—this is the case across respondents from different roles, company sizes, and geographies. This tally doesn't take into account the full number of bespoke tools they build for themselves or internally built tools their company provides. We saw a major rise in these "internal" company tools—covered in more detail below. **The 2026 AI toolstack (Category: General AI)** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply.* | Claude | 78% | | --- | --- | | ChatGPT | 65% | | Gemini | 48% | | Company internal tools | 29% | | Perplexity | 13% | | Chart title: The 2026 AI toolstack (Category: Design-specific) Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply. | Chart title: The 2026 AI toolstack (Category: Design-specific) Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply. | | Figma AI / Make | 68% | | Image generation (e.g. Midjourney, Weavy) | 54% | | Company internal tools | 18% | | Video generation (e.g., Runway, Sora) | 17% | | Adobe AI | 16% | | Chart title: The 2026 AI toolstack (Category: Coding and dev) Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply. | Chart title: The 2026 AI toolstack (Category: Coding and dev) Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply. | | Claude Code | 65% | | Cursor | 39% | | Lovable | 19% | | v0 by Vercel | 17% | | OpenAI Codex | 15% | | Company internal tools | 12% | | GitHub Copilot | 9% | | Replit | 8% | | Bolt | 2% | | Chart title: The 2026 AI toolstack (Category: Research) Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply. | Chart title: The 2026 AI toolstack (Category: Research) Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply. | | AI notetakers (e.g., Otter, Fathom) | 41% | | AI analysis tools (e.g., NotebookLM, Looppanel) | 31% | | General research tools (e.g., Dovetail, Maze) | 25% | | Company internal tools | 18% | | Analytics with AI features (e.g., Hotjar, Fullstory) | 17% | Coding tools have found their place in the go-to design stack. In 2026, 76% of all respondents say they've used an AI coding tool like Claude Code, OpenAI Codex, Cursor, or GitHub Copilot. 85% have used these and/or an app builder like Lovable, Replit, or Bolt. ### Claude has overtaken ChatGPT as the primary general AI tool 78% of respondents use Claude, compared to 65% for ChatGPT, which led in 2025. Gemini comes in third at 48%, and Perplexity has dropped from 34% to 13%. 65% of overall respondents are using Claude Code, which didn't even make it onto our 2025 survey because it hadn't launched to the public yet. Disclosure: Anthropic, the creator of Claude, is a partner of the AiiD26 survey. The survey was conducted independently of our partners. **General AI tools: year-over-year shift** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply. Values compare 2026 to 2025.* | Tool | 2025 | 2026 | | --- | --- | --- | | Claude | 52% | 78% | | ChatGPT | 88% | 65% | | Gemini | 21% | 48% | | Perplexity | 34% | 13% | ### The canvas is shifting According to the UX Tools [State of Prototyping](https://survey.uxtools.co/spring-2026) survey, Figma remains the most-used design tool in 2026. But how designers use it may be changing. Some say Figma is their starting point for AI. They explore multiple directions in the canvas before going deep in code, or they create wireframes in Figma and pass them to AI. It remains a favorite place to ideate and collect inspiration, and respondents also pointed out that it's still the best tool for team collaboration. Other designers are using Figma as a "scalpel" or finishing tool in their AI-first design process. They prototype in a dev environment, then flip back to Figma for high-fidelity execution or precise tuning like perfecting radii and padding. On the other end of the spectrum, we heard from code-first designers who rarely spend their time in Figma now. At Maze, designers work in Cursor from day zero. At Watershed, the design team has fully transitioned to working directly in the codebase. > "Figma has shifted from being the primary design tool to a canvas for quick exploration and polishing details as input for Claude Code." > > — Phil Vander Broek, Head of Design, AI, Superhuman > "Traditional prototyping took two weeks on average. Now, using vibe-coded prototypes, it's completed in hours. We use Figma as a scalpel now—pull an area out for specific cuts and push it back into the coded prototype." > > — Mark Boyes-Smith, Head of AI Design, Miro > "Fine grain control is the biggest gap. It's the tuning. It's the 'I know what I want in my head, but you're just not giving me what I want.' So I need to go into Figma and tell you what I want and then put that back." > > — Shali Nguyen, Head of Consumer Experience Design, DoorDash Some designers are taking a middle path by using a newer generation of design tools like [Paper](https://paper.design/), which are rebuilding the canvas in HTML and CSS. This allows them to continue using a canvas that connects directly to agentic tools, alongside other inputs like text prompts or code. This tool adaptability is part of a broader AI-enabled shift that allows designers to customize their workflows—whether they want to start with code, mood boards, words, pen and paper, or an AI tool's brainstormed idea. Ryan Mather, a designer at Anthropic, says, "The other week I painted an interface and shared it as input to my AI tool of choice!" ### Late entry: Claude Design Just after we closed our 2026 survey, Anthropic [launched](https://www.anthropic.com/news/claude-design-anthropic-labs) Claude Design, an AI design workspace that generates complete prototypes, decks, one-pagers, and marketing assets. It's accessible to non-designers like founders, PMs, and marketers who don't live in Figma the same way designers do. We're interested to see how this tool's adoption will evolve and whether it becomes part of the standard designer's toolkit in the next report. **WHAT MAKES TOOLS STICK** # 3. Output quality is both the most popular driver of stickiness and the biggest weakness in current AI design tools 80% of designers say reliable, high-quality output is what makes an AI tool stick. At the same time, 62% cite inconsistent or unreliable output as their biggest challenge when using AI for design work. We heard designers say the tools just "aren't quite there" yet. But they also express how life-changing it would be if they were. Many believe that models will improve so rapidly that AI designs will soon pass art school standards and rival human craft. Other major barriers to adoption include steep learning curves and poor integration with existing tools. At the company level, security and compliance concerns—along with budget constraints—continue to limit adoption, which may contribute to the proliferation of internally built design tools (see examples below). **What makes an AI tool stick** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply.* | Reliably high-quality output | 80% | | --- | --- | | Easy to use and well-designed | 61% | | Fits into how we already work | 57% | | Integrates with existing tools | 54% | | Offers specific capability no other tool provides | 35% | | Trusted recommendation | 30% | | My team or company standardized on it | 24% | **Top challenges of using AI in design work** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply.* | Unreliable or inconsistent output quality | 62% | | --- | --- | | Lack of control over output | 45% | | Lack of product/brand context | 42% | | Hard to integrate with existing tools | 39% | | Security, compliance, legal concerns | 37% | | Too time-consuming to learn | 33% | | Concern about skill atrophy | 28% | | Budget or cost constraints | 28% | | No significant challenges | 5% | Because their attention is spread across an ever-growing set of tools, designers are craving both a single, integrated platform that spans the full design process (in the vein of Figma or Sketch) and best-in-class tools for each step. They want seamless integration with existing systems and workflows. And they want to know that these tools will remain stable over time. In an environment where underlying models are evolving so quickly, that expectation may be difficult to meet. > "We need a tool that allows us to seamlessly go from research to discovery to concepting and sketching to prototyping to high fidelity to code, and not necessarily in a linear fashion." > > — Individual contributor, Growth-stage company > "One go-to tool or less jumping between tools for an optimal end result." > > — Executive, Publicly traded company > "[I'd like] one major stack that can be deployed without worry it will be replaced in a few months." > > — Executive, Growth-stage company **DESIGNERS AS TOOLMAKERS** # 4. Designers are building their own tools Designers are building bespoke tools—first for themselves, then for their teammates—by encoding their judgment, taste, and design systems into workflow infrastructure. This is one of the most compelling shifts we've observed over the last year: designers as toolmakers. The most sought-after designers today share this behavior. If designers were previously evaluated on their output, now they are evaluated on both their output and the workflows they build. ### Personalized software, custom-made for how you work At the individual level, designers are creating microtools when they have a specific need that an off-the-shelf tool can't meet, like automating a tiny but repetitive step in their process. They're also generating full-fledged apps that better fit their styles of thinking and communicating. "You can mold the tool to the thing you're doing," says Ryan Mather, Designer at Anthropic. "Need a quick dark mode simulator? Sure, just ask for one. Need to mock up mobile, wait no, tablet, wait no, desktop? Sure! Not a problem to make these schleps that used to be so hard, sometimes you just wouldn't make them." Examples of tools created by designers: Gavin Potenza built [Moodboard 3000](https://www.linkedin.com/posts/gavinpotenza_i-built-a-figma-plugin-moodboard-3000-activity-7449862581902618625-mQa9), a Figma plugin that generates composed moodboards. Amelia Wattenberger built a [thought-organizing tool](https://wattenberger.com/thoughts/our-interfaces-have-lost-their-senses/) that turns spoken or typed ideas into structured cards in real time. Brian Lovin built [shiori.sh](https://www.shiori.sh/), a minimal read-it-later desktop app that captures content, transcripts, and summaries. One survey respondent noted: "I built an app that adds an iPhone bezel when I 'copy as png' from Figma and it's so magical!" ### What designers are building for their teams At the company level, designers are building shared tools that solve problems for other designers as well as for engineering, PM, and marketing. The most AI-forward teams we spoke with—especially those at enterprises—have invested in infrastructure and promotional channels that make it easy for designers to evangelize and borrow home-grown tools. Examples include show-and-tells, internal marketplaces, and skill libraries. Read more in Teams. > "Our team built ProtoDash, an AI-powered product playground with Stripe's design system baked in. Now anyone can build a realistic prototype in minutes. And to scale content quality, we built Dante, a content tool integrated in the workflow—Slack, GitHub, docs, and the CLI—that makes it easy to do the right thing." > > — Katie Dill, Head of Design, Stripe > "I can enforce a high quality bar on everything that gets built, even if it's not coming over my desk directly. I build the design system in the actual app now—buttons, dropdowns, popovers—and that's a point of leverage to make sure even things I don't touch look good." > > — Nick Inzucchi, Product Designer, Cursor > "We have an org-wide skill library in Claude, where I installed a visual brand skill that helps everyone across AirOps create on-brand landing pages, data visualizations, slides, and more. It caught on like wildfire. These would normally take my small brand team days to produce. It ended up inspiring an AirOps feature for our customers." > > — Jessica Rosenberg, Head of Brand, AirOps Inside Anthropic's internal design toolkit Anthropic's design team uses a set of internal tools that speed up their work. A system of microtools—each supporting different phases of the design process—can become an extremely powerful asset for your team. Ideation sandbox: Takes a brief and generates many UI directions for early-stage exploration Design system picker: Plugs Anthropic's fonts, colors, and components into Claude so prototypes start on-brand Research index: Makes Anthropic's user studies queryable, so anyone can check what's already known in the organization Looping PRs: Once a change is approved, an agent opens the PR in GitHub and watches it through CI (continuous integration) until it merges—/loop and done Content guardrails: A Slack agent that scans production code for copy that drifts off-brand and suggests rewrites ### Internal tool usage scales with company size 74% of designers at enterprise companies (2,000+ employees) use internal tools, compared to just 26% at small organizations (up to 50 employees)—a 5x difference. At enterprises, these tools rank as the second-most-used general AI tool, behind Claude. **Designers using internally-built AI tools, by company size** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply.* | 1-50 employees | 26% | | --- | --- | | 51-500 employees | 34% | | 501-2,000 employees | 48% | | 2,000+ employees | 74% | Here are a few examples of how internal tools spread: Internal open-sourcing: Anthropic has a separate GitHub repo that open-sources what team members build, making it easy to borrow from one anothers' projects and adapt them into their own bespoke tools. Playgrounds: Notion built a prototype playground—an internal, shared Next.js codebase where designers can create interactive product prototypes. Each designer works in their own folder, with a shared set of Notion-like design system components already available out of the box. Leaderboards: Stripe created an internal ranking of the most popular employee-built plugins and a spreadsheet of agents anyone can run. **CONFIDENCE VS. NOISE** # 5. Confidence in tools is up, but so is the noise Designers are feeling overall more confident in their AI toolstack compared to a year ago, but many are constantly exploring new options. Almost half of designers say they're still searching for their go-to tools, while 37% say they've settled on a clear set of tools for most workflows. On average, designers who work at organizations with strong support for AI adoption report more confidence around which AI tool to use for which design tasks. **Designers' confidence in knowing which AI tools to use for which design tasks, compared to a year ago** *Question type: Single-select; percentages sum to ~100% because respondents could select only one option.* | Much more confident—I have clear go-to tools for most workflows now | 37% | | --- | --- | | Somewhat more confident—I've figured out a few, still searching for others | 49% | | About the same | 6% | | Less confident—the landscape feels more overwhelming than it did a year ago | 6% | | I wasn't using AI tools a year ago | 2% | ### Keeping up is taxing The tools seem to change daily as capabilities evolve. Some designers describe this as a never-ending "molting process" that requires rethinking their entire workflow regularly. Many others report a growing workload, both to keep up with testing the latest tools and because expanded capabilities give them more possibilities of what to work on. Research from Harvard Business Review [corroborates](https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it) this: AI doesn't reduce work, it intensifies it. > "Every day someone is sharing a tool/workflow they saw. It's both exciting and draining, as it creates a lot of noise that makes it harder and harder to understand what tools are worth considering. Workflows and process cannot change every day, it becomes unsustainable." > > — Individual contributor, Growth-stage company > "Many companies are missing out on great talent because they think a candidate needs to come in already knowing every new AI tool. These technologies change constantly—you can't expect mastery yet. And in reality, founders themselves are often still figuring out the landscape. That can lead to over-indexing on tool coverage in hiring as a way to compensate. The better approach is to hire for potential, taste, and motivation—and to invest time in understanding how designers work and what drives them." > > — Garrett Fowler, Design Recruiter and Founder, Offsite ## Where do we go from here? We're curious how design tools and workflows will continue to evolve as more work becomes mediated by AI. What will craft look like in the age of AI? Visual output has been the dominant proof of craft, but this could become more difficult to discern as AI capabilities improve. Will the bar for craft shift to storytelling and brand? How will designers use Figma going forward? Figma has been at the center of design workflows for over a decade, defining how teams collaborate and build together. But as design tools become more connected to code and agents, what will its role become? Will we see a consolidated design platform? Designers are stitching together many more tools to get work done than they did a year ago, and in some cases they're building their own. But many are also asking for the opposite: fewer tools, better integrated, and in one place. Will a single platform emerge as the default, or will orchestration across multiple tools become the norm? Will we ever have a "go-to design stack" again? If designers can build their ideal software or pick from a deluge of new off-the-shelf tools, we wonder whether every team, or perhaps every person, will have a different default stack—and whether even that will stay consistent for more than a few months at a time. How long will it take for model capabilities to catch up to what a human designer can do? If it's true that AI's quality will mature to handle most production and visual design work, what timeline are we talking about, and what parts of design will remain out of reach for AI tools? ## Key takeaways Designers are using double the number of off-the-shelf AI tools than they did a year ago. Nearly all our respondents use AI in their design work weekly or daily. Most are still figuring out their go-to tools, and we expect the toolstack to become fluid. Designers are building their own design tools with AI. They're inventing microtools and robust agentic workflows to automate manual work, create the design software they've always wished they had, or simply save cash. Enterprises are solving for security and compliance by building internal tools. Designers at larger companies are much more likely to use internally built AI tools. This workaround allows enterprise design teams to adopt AI and stay on the forefront. Claude appears to have overtaken ChatGPT in the last year. Claude and Claude Code are the most commonly used commercial tools in the "General AI" and "Coding" categories. We're still in an era of unreliable output quality. The landscape of AI design software in early 2026 is orders of magnitude more robust compared to a year ago, but designers still cite inconsistent output quality as the top challenge of AI tools in design work. # Further reading "[What's next for design tools](https://www.proofofconcept.pub/p/design-toolings-future)" — David Hoang "[State of Prototyping: Spring 2026](https://survey.uxtools.co/spring-2026)" — UX Tools [Designtools.fyi](https://designtools.fyi/) "[AI Doesn't Reduce Work—It Intensifies It](https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it)" — Harvard Business Review "[How I AI: Stripe's Owen Williams on Killing 'Blurple Slop' with an Internal Prototyping Studio](https://www.chatprd.ai/how-i-ai/stripe-owen-williams-on-buildling-internal-prototyping-studio)" — Owen Williams and Claire Vo # Craft 02 Craft Craft in an age of infinite output Changing what we make and how we make it As designers keep pace with faster-moving product teams, they're using AI to accelerate tasks across the design process, from discovery and ideation to prototyping and frontend polish. They're also stretching the scope of their role. Just a year ago, designing with code was typically limited to design engineers or highly technical designers. But in this year's study, half of our respondents, across product and brand design specialties, say they've shipped AI-generated code to production. Many say their teams expect a new design deliverable: fully functional prototypes. Whether code becomes a primary medium or a tool designers use selectively, this shift toward building appears to be energizing. Designers using AI-assisted coding and prototyping are more likely to report feeling more creative and capable at work as a result of AI. At the same time, respondents worry that speed can cut into the incubation time needed for strong ideas, and that outsourcing parts of the process may limit how younger designers develop judgment. Most still place greater trust in human design skills than in AI. Coding Prototypes Speed vs. quality Human skills vs. AI Joy at work Where do we go from here? Key takeaways Further reading - **76%** — have used an AI coding tool - **50%** — have shipped code to production - 80% — prefer human judgment over AI — in areas like creative direction **CODING** ## 1. Half of surveyed designers have shipped code Company typeEarly-stage Growth-stage Publicly tradedSource: AI in Design Survey, Q1 2026Borrowed this number placement from 2025 report, this isn't prescriptive :)Use #'s in the table plsHave you shipped AI-generated code to production?AiiDNote - using AiiD attribution to simplify DF + Foundation Capital 50% of all respondents have pushed AI-generated code to production, including both product and brand designers. Only 20% of respondents identify as design engineers, underscoring how quickly coding capabilities have spread beyond traditionally technical roles. 76% of respondents say they've used an AI coding tool that augments developers, like Claude Code, Cursor, OpenAI Codex, and GitHub Copilot. If we include AI app builders like Lovable and Replit, that jumps to 85%. Early-stage designers are more likely to ship code at 68%, compared to 33% at publicly traded companies. **Designers who have shipped AI-generated code to production, by company stage** *Question type: Single-select (yes/no); percentages show the share within each group who answered yes.* | Early-stage | 68% | | --- | --- | | Growth-stage | 50% | | Publicly traded | 33% | ### Designer founders lead in shipping code Founders are the shippy-est group (70%) compared to ICs, managers, and other executives. Some of the strongest designers we know have left design teams entirely in favor of [starting their own companies](https://designerfounders.substack.com/). We think designers are uniquely wired for this path, with a deep sensitivity to users, a bias toward making ideas tangible, and a high bar for quality. Those abilities—combined with AI tools—make it possible to go from idea to prototype to fully shipped product with far less engineering support than before. We're also seeing that as many leaders as ICs have pushed code. With AI, they're able to get in the weeds to quickly prototype an idea or drive a project forward. Nearly half of executives we surveyed are shipping, and in turn, they're challenging what's expected of leaders. Read more in Teams. **Designers who have shipped AI-generated code to production, by role type** *Question type: Single-select (yes/no); percentages show the share within each group who answered yes.* | Founders | 70% | | --- | --- | | Executives and Managers | 47% | | ICs | 47% | ### What are designers shipping? All types of code. They're taking over frontend polish, building backend systems, integrating design system architecture, spinning up beta products, adding motion, and more. > "Designers can directly fix code issues. No more 'blocked on engineer' for front-end tweaks. Design system components—buttons, drop-down components, popovers—are auto-integrated in Cursor." > > — David Stinnette, Director of Product Design, Samsara > "Design QA has significantly changed. Previously, we would just create tickets and hand them off for the eng team to fix. But now, we can go in and address these smaller tickets (copy, spacing, change component, etc...) ourselves." > > — Individual contributor, Growth-stage company > "Using skills in Cursor, I have been able to integrate good motion design into my work." > > — Individual contributor, Early-stage startup > "Localization has become almost completely automated—from text phrasing and translation to pull requests and merges." > > — Individual contributor, Publicly traded company A designer who can bridge design and code is especially valuable on small teams, where that versatility accelerates shipping. But this isn't limited to early-stage companies. As roles merge (more on that in Teams), the expectation for technical fluency is rising across org sizes—without losing what designers uniquely bring: a strong point of view on what to build and how it should work. ### Automating code review for velocity Across the companies we surveyed or spoke to, we heard a range of practices around code review or handoff. On the high-speed side, companies are using AI to assess the risk level of push requests (PRs) initiated by designers and automatically push low-risk changes. At other companies, designers are deploying to separate sandboxes for engineering review, or handing off prototypes for engineers to rebuild in production code. At DoorDash, the PR process has stayed the same, says Shali Nguyen, their Head of Consumer Experience Design. "Everything goes through GitHub PR approvals, and designers are held to the same production-quality bar as our engineering counterparts. We've added a design review step where designers are tagged to review and approve UI changes before they go to production." > "All of our designers have submitted production PRs. They're also partnering with engineers on their teams to review code and build their technical skills." > > — Hannah Hudson, Head of Design, Watershed > "I'm floored by my literal ability to ship PRs ... We do post-facto PR reviews with an AI risk scoring system and reviews from multiple AI bots, so I can also merge my own PRs with no oversight in many cases. I fix user issues while we're discussing them and ship before the meeting ends." > > — Brooks Solveig, Staff Product Designer, Flux With so many more designers cultivating coding capabilities, what will happen to the "design engineer" title? We explore these themes of blurring roles in Teams. **PROTOTYPING** ## 2. Working prototypes are now shipping at every stage of the design process - **43%** — say their companies expect working prototypes - **36%** — say projects now start with working prototypes In our 2025 survey, we reported that prototyping was emerging as a key use for AI, and that designers were beginning to skip static mockups to go straight to vibe coding prototypes. This trend has continued. Now, prototypes are an expected design output for 43% of respondents. Some designers now do all their exploration work via coded prototypes rather than in a canvas, or they simply incorporate prototyping into a broader process that still includes mockups in Figma. Prototypes fill a gap in the classic set of design deliverables, because unlike mockups, they make it easy to evaluate states in a user flow. Designers we spoke with say that prototyping helps circulate ideas and align with engineering. Others expressed concern that prolific prototyping invites too many options—and that these options often look alluring enough to miss their poor design. > "It could take two weeks to create a prototype, but it's not the thinking that takes two weeks. It's the effort of using tools to manually craft all of the details required. And if someone comes along and says 'you could have done that five degrees to the left,' it's difficult to adjust that experience. It's like a fired pot, you have to smash it and start again. With vibe-coded prototypes, the solution is malleable, takes a few hours to produce, and you benefit from all that wonderful collaborative energy." > > — Mark Boyes-Smith, Head of AI Design, Miro > "We're building prototypes in the code base—in the actual product experience, with real customer data, so that designers can see how their design will perform in the context of an account. This is helpful to identify edge cases and see how the data flexes inside the designs they're creating." > > — David Stinnette, Director of Product Design, Samsara > "Most of our design critiques have shifted to showcasing prototypes and looking at the user journey or story we're telling. Previously we'd zip around Figma looking at static screens and miss the connective tissue of the workflows. Less handoff meetings where we're throwing things over the fence ... designers invite engineers and PMs into Builder/Lovable or publish Cursor to an internal site and get feedback live instead of handing over lossy static assets." > > — Executive, Publicly traded company > A closer look Making prototyping a team-wide capability At Anthropic, designer Nate Parrott built an internal tool where Claude generates interactive prototypes using the company's full design system. Designers use it to create features and explore animation options; non-designers like educators and salespeople use it to create visual artifacts they couldn't make before. "It's as if they were a designer all along and they were just blocked on this one technical skill," he says. Prototypes are shareable with a link, and feedback goes directly to Claude for revision. This internal tool became a precursor for Claude Design. > A slide deck created in the tool is being edited, with feedback sent directly to Claude for revision. **SPEED VS. QUALITY** ## 3. Some say AI's speed removes design bottlenecks, others worry about cutting corners Designers are faster with AI. But are they better? And who sets the quality bar when everyone—including PMs and engineers—can produce plausible-looking interfaces? When we asked respondents about the impact of AI on various aspects of their design work, they said their speed, efficiency, and creative range have benefited the most. **AI's impact on aspects of designers' work** *Question type: Single-select per row; respondents could also select "No change" or "Mixed," which are omitted here. Improved and Decreased do not sum to 100% by design.* | Aspects of designers' work | Improved | Decreased | | --- | --- | --- | | Speed and efficiency | 85% | 2% | | Creative range and exploration | 65% | 7% | | Personal job satisfaction | 53% | 18% | | Output quality | 51% | 8% | | Confidence in design decisions | 46% | 6% | | Sense of ownership over your work | 32% | 23% | | Collaboration with teammates | 29% | 20% | But when we asked survey respondents what they'd miss most if they lost access to AI tomorrow, they were more likely to say they'd miss the ability to build and the breadth of creative expression. Designers have told us they're now more technically empowered to: Implement more frontend polish Try more complex designs Get up to speed on a new industry Improve their accessibility knowledge Synthesize disparate data sources to make informed decisions Complete a rebrand in a week And for resource-strapped companies in particular, AI has unlocked new levels of capability and quality that were too expensive to reach in the past. ### Keeping pace Now that product teams can build features in hours instead of weeks, designers are feeling even more outnumbered by PMs and engineers—and are stretched as they provide quality design guardrails for their prolific teams. So what are designers doing about it? Encoding quality into tools: They're finding creative ways to enable a quality bar for teammates that have stepped into designer territory (40% of respondents say their PMs and engineers are doing more design work—read more in Teams). Designers are preprogramming design system components and brand guidelines into coding tools so that anyone producing interfaces starts from a shared quality baseline. Prioritizing differently: They're rethinking how to allocate their time to projects that would most benefit from deep design thinking—like ambiguous UX problems or connecting with user needs. Moving from one design process to many: Designers are approaching each project differently, considering what type of process to use depending on timeline and output needs. > "We really do think more about polish because we can deal with implementing it in a fairly low-cost way. Having opinionated taste is no longer a luxury that we need to balance." > > — Individual contributor, Growth-stage company > "Since engineering capacity has skyrocketed, we're forced to think about how to do more with less. How do we put guardrails on product experiences that meet a threshold of quality? How do we identify ambiguous problems that need UX thinking versus "commodity" features that have plenty of precedence and don't need differentiation?" > > — Manager/Lead, Publicly traded company ### Does speed support more exploration or less? Many designers told us AI helps them iterate faster to match engineering's pace. But not without cost—we also heard that code forces commitment too early and that speed compresses the open-ended phase where designers develop their judgment. > "Once a problem is identified, you can generate a million ways you could potentially solve that problem to see if any of them are even remotely good." > > — Hannah Hudson, Head of Design, Watershed > "Code forces you to commit to your first idea and go deep—at the expense of the broad exploration that Figma made easy. The breadth of exploration is the limitation I feel most." > > — Nick Inzucchi, Product Designer, Cursor > "[AI] generates ideas so much faster than me. It is also more willing to generate bad ideas than my team." > > — Executive, Publicly traded company > "It's very easy to put my prompt in, get an answer back, do a quick shuffle, and think: This is great. Job done. We don't really want designers to do that. We want people to have radically divergent concepts, so we can invest our time debating the space in the middle." > > — Mark Boyes-Smith, Head of AI Design, Miro > "We used to sit together and brainstorm user flows and app flows from scratch ... manually exploring multiple possibilities, eliminating each step by step. Now, our ritual has shifted, and we start by validating and evolving explorations proposed by AI instead of beginning from a blank slate." > > — Individual contributor, Agency > "To me, design is the planning stage and code is the implementation stage. I don't like mixing those two because you have different goals in those workflows. The planning goal is finding clarity on what you're doing or what you want to build. Execution is about making it work well. If you try to combine them, it's not ideal for either." > > — Karri Saarinen, Co-founder and CEO, Linear As product development speeds up with AI, design teams that don't keep up may risk falling behind or losing influence. They're balancing this with the cost of letting go of crucial parts of the design process, and they're finding middle ground by learning to work differently with their counterparts, which we explore in Teams. Joel Lewenstein, Head of Product Design at Anthropic, says, "There's a socially interesting dimension, which is that if a feature can now be built in 48 hours, asking for a week-long classic design sprint to explore divergently is a huge ask. "At Anthropic, we've started to compress design exploration with new internal tools that generate many UI variations, index user research across the company, and enable rapid dogfooding. But there's still an ineffable, slow part of exploration time that we need to protect—going for a walk, letting the brain relax, letting the neurons connect—and that is a work in progress." **HUMAN SKILLS VS. AI** ## 4. Designers value their taste, judgment, and user understanding over AI For about 80% of respondents, AI assists but doesn't replace their own quality meter, final visual polish skills, and creative direction. 60-70% of designers also reported that they still rely on their judgment for understanding user needs and context, defining the problem, storytelling, systems thinking, and design architecture. **Where designers most rely on their own craft and judgment (over AI)** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply.* | Final visual polish and quality decisions | 81% | | --- | --- | | Creative direction and aesthetic judgment | 79% | | Understanding user needs and context | 71% | | Strategic framing and problem definition | 67% | | Storytelling and communicating design rationale | 65% | | Cross-functional communication and alignment | 62% | | Systems thinking and design architecture | 61% | | Ethical and inclusive design decisions | 54% | > "It's like a burrito that took a minute and a half to microwave. I am happy with it, because I am in a rush and I am thrilled it's ready so fast. But is it actually good? Heck no. This same phenomenon is what leads to a ton of slop. It's easy to get enamored by a decent one shot and overlook the flaws. A good practice is to ask yourself, would I be happy with this if it took me a week to make this? If the answer is no, it's probably worth some further prompting and prodding to make it something you're truly proud of." > > — Katie Dill, Head of Design, Stripe > "Without the weight of tech debt, brand/interface value, and the complexity that comes with mature, post-PMF product, it's easy to claim 'Figma is dead,' and go all-in on prototypes and gen-AI front-end. But at scale, or when dealing with mature problems, I've found that there's still no substitute for human innovation in UI design, and an enduring need for a design environment where we own the pixel-level choices, and this has been consistent when speaking with other leaders in similar positions." > > — Executive, Growth-stage company ### Ownership of AI output Only 9% of respondents feel that it's genuinely difficult to separate AI's input from theirs. Many more feel they have full ownership (40%) over AI-produced work—that the direction and judgment still comes from them—or that it's "mostly" theirs (43%). **How designers feel about ownership of the AI-assisted work they produce** *Question type: Single-select; percentages sum to ~100% because respondents could select only one option.* | Full ownership: The direction and judgment are mine, AI just executes | 40% | | --- | --- | | Mostly mine: But I'm aware AI shaped some outcomes I didn't fully control | 43% | | Shared: It's genuinely hard to separate my contribution from the tool's | 9% | | Uncertain: I'm still working out how I think about this | 8% | Ryan Mather, a designer at Anthropic, says, "AI is powerful because the tool itself can think and experiment. It's up to the person holding the tool to decide where to direct that. Every good designer has a point of view, and AI is like a point-of-view amplifier." **JOY AT WORK** ## 5. Designers who build are feeling more creative and capable We were curious how designers feel about their work as AI usage grows. 53% of overall respondents report a positively transformed relationship with their work as a result of AI. 18% say it has decreased their work satisfaction. **How AI has affected designers' job satisfaction** *Question type: Single-select; percentages sum to ~100% across the rating scale because respondents could select only one option.* | Change in personal job satisfaction | Percent of respondents who selected | | --- | --- | | Significantly improved | 20% | | Somewhat improved | 33% | | About the same | 26% | | Somewhat decreased | 12% | | Significantly decreased | 6% | | Mix of improved and decreased | 3% | This trend was even stronger for designers who "build." We noticed that designers who ship code or create prototypes with AI were twice as likely to feel more creative and capable at work—while also feeling that the quality bar is higher. The Spring 2026 UX Tools [State of Prototyping](https://survey.uxtools.co/spring-2026) study echoes this finding. **Compared to before AI, designers now feel...** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply. Compares designers who code & prototype with AI vs. those who don't. Ratio shows the magnitude difference between the two groups.* | | Designers who code & prototype | Designers who don't | Difference between designer groups | | --- | --- | --- | --- | | More creative and capable | 65% | 35% | 2.5x | | Much more confident in their tool stack | 57% | 24% | 2x | | That they face a higher quality bar | 27% | 17% | 2x | More joy at work > "What excites me the most is pushing a higher product quality bar via my own PRs and having the tools to explore and debate scope and feasibility (so I can push for more complete UX flows, accessibility, component improvements)." > > — Individual contributor, Growth-stage company > "I'm having way more fun with the work. I can spontaneously follow my curiosity into new domains like frontend code and shaders, instead of getting blocked by what I haven't had time to learn yet." > > — Jason Marder, Independent designer > "The tools were more creative than anticipated. With Claude Code and Cursor, I've been surprised by how they can riff off ideas I have and show me things I didn't know were possible with HTML/CSS/JavaScript." > > — Founder, Educational organization > "Unexpectedly useful is just being able to ask questions about the codebase—how things work, what's the structure, where the data comes from, what component is being used, etc. Things that I used to need to ask an engineer." > > — Individual contributor, Growth-stage company > "As a designer with more than twelve years of experience ... The scope of what I can do, both inside my company and outside of it, feels bigger than before. I genuinely think this is one of the most exciting times in design history to be a designer." > > — Individual contributor, Growth-stage company > "Honestly when it works it makes work fun. I've been able to lean into more exec and product-level things with Claude." > > — Executive, Growth-stage company But these findings don't tell the whole story. Respondents also describe new stressors—like the pressure of speed, the need to "always be prompting," and the fear that the human-centered skills that have always made design "design" will be less valued than technical skills. There are also increasing feelings of isolation. Designers are working solo more often since 2025, with 4x more respondents saying that collaboration has decreased (20% vs. 5% in 2025). We explore this more in Teams. Less joy at work > "AI is great, but the expectation of immense speed or 100x productivity has annihilated the joy and pride of the job." > > — Individual contributor, Early-stage startup > "In the past, there was a clear ROI to learning a new tool. Now it feels like a grind because every week there seems to be a new major update. I put in that investment and then a couple months later the tools change. The dividends don't pay out as much." > > — Alexander Cheung, Senior Product Designer, Pinterest > "There's loneliness replacing the collaborative energy. Waiting for AI to process replaces flow state. We can do cool things now, but with everyone building independently in a terminal, it's devoid of the interaction that we often need to feel fulfilled." > > — David Stinnette, Director of Product Design, Samsara > "The question is almost never 'Should we use AI here?' but 'Can you try to use AI more? Where else can we use it?' Almost as if we need to try to infuse it as much as possible in order to impress." > > — Individual contributor, Growth-stage company > "We all worry about craft atrophy and how successful younger associates ... will be given that AI has been more core to their journey than those of us who did it all by hand; therefore, will they have enough wisdom in decisions vs. leaning on AI to outsource the thinking." > > — Executive, Publicly traded company > "It seems like the hard skills of AI tooling are valued more than the soft skills that paved the way for user experience in the first place ... hiring managers only value shipped work." > > — Individual contributor, Publicly traded company ## Where do we go from here? As model capability accelerates, we'll be tracking: What does it mean to design for agents—and to design for humans working alongside them? This was a topic that surfaced surprisingly little in our conversations and survey responses. The old interaction patterns won't apply when we make software for agents; they want access to data and clarity. As Alexander Cheung, Senior Product Designer at Pinterest, puts it, "If AI is designing for AI, it doesn't even need to be delightful. It just needs to work." But we can also think of agents as a new archetype of coworker. Even if we're no longer moving pixels ourselves, we still want to design alongside AI, so we need better ways to communicate with it. How will designers think about accountability? Agents aren't accountable for mistakes. When a self-driving car gets into an accident, a human needs to answer for it. The same goes for agent-produced design work. Humans need to be able to control AI tools, and perhaps it's this element of accountability that will keep humans in the process as AI takes on more production work. Does all this velocity produce better design, or just more of it? Faster shipping is one of the most visible changes from 2025 to 2026. Whether the floor of quality is rising is still debated. Will humans still have an edge in taste? AI may close the gap on craft and usability faster than on taste, judgment, and cultural foresight. Humans have a unique advantage in our ability to tell compelling stories that resonate with our fellow humans and to curate what AI should and should not do. But how long will that gap stay meaningful, and will companies keep paying a premium for it? What do designers need to learn to position themselves for multiple possible futures? ## Key takeaways Coding is part of more designers' workflows. Half the designers we surveyed across brand and product design, including founders and executives, have shipped AI-generated code to production. Coding is a new surface to collaborate with engineers. Prototypes are now a key design output. Continuing a trend we saw in 2025, design teams are prototyping more, and in some cases skipping static mockups. Designers seem to value the creative range that AI gives them. They're also benefiting from more speed, but this creates mixed feelings. Some say it unlocks higher quality work, while others believe they're cutting corners on the design process. Designers aren't handing the keys over to AI completely. Over 80% of those we surveyed still rely on their own judgment around craft, quality, creative direction, and more. People are feeling a mixture of anxiety and joy. There are serious worries about craft atrophy, loneliness, and the hopelessness of trying to keep up with new technology. At the same time, most designers we surveyed reported more overall job satisfaction as a result of AI. ## Further reading "[Output Isn't Design](https://linear.app/now/output-isn-t-design)" — Karri Saarinen "[Beyond Prototyping](https://www.aidesignfieldguide.com/articles/nate-parrott)" — Nate Parrott "[The design process is dead. Here's what's replacing it.](https://www.lennysnewsletter.com/p/the-design-process-is-dead)" — Jenny Wen and Lenny Rachitsky "[The 2026 AI Design Field Report (tools, process, and what's working)](https://www.youtube.com/watch?v=Y0n6F9VlLVc)" — Steven Haney and Michael Riddering "[Your product has a new user. It's not human.](https://www.elenaverna.com/p/your-product-has-a-new-user-its-not)" — Elena Verna "[Design in Tech Report](https://designintech.report/)" — John Maeda # Teams 03 Teams Redesigning the design team As designers do more—and more people do design—company norms fall behind Companies are putting more budget, programs, and encouragement behind AI adoption than they were a year ago, but designers are still mostly teaching themselves and learning from one another. Leaders can enable this grassroots learning by creating space for tinkering and experimentation. Some designers are now doing more PM and engineering work, and vice versa. They're also becoming architects of AI workflows and builders of custom tools. But despite rising expectations for both volume and quality of work, few companies have updated their official policies around performance reviews, team structure, and hiring practices. Most design leaders we surveyed expect to maintain or grow their design headcount, and some are shifting resources to AI-native hybrid design roles. Hiring managers are looking for AI fluency, but they still hold a high bar for craft, vision, strategic thinking, and storytelling. AI adoption Role changes Expectations & structure Hiring & careers Where do we go from here? Key takeaways Further reading - 87% — of designers surveyed receive moderate or strong support — for AI adoption from their companies - **28%** — of leaders have changed formal performance review metrics, incentives, career ladders, and hiring processes - **60%** — of leaders expect to keep or grow design headcount **AI ADOPTION** ## 1. Companies are investing more in AI adoption, and designers still learn the most from one another Companies have stepped up their support for designers adopting AI. 87% of all respondents report at least moderate organizational support for AI adoption, and 53% say that support is strong—via budget, active encouragement, and formal programs. **Support for AI adoption by company type** *Question type: Single-select; each column sums to ~100% because respondents at each company type could select only one option.* | Support for AI adoption designers receive | Company type | Company type | Company type | Company type | Company type | | --- | --- | --- | --- | --- | --- | | | Early-stage | Growth-stage | Publicly traded | Consultancy or design agency | Other | | Strong support—dedicated budget, active encouragement, and formal programs | 64% | 59% | 53% | 34% | 31% | | Moderate support—some tools provided, limited formal structure | 23% | 34% | 35% | 47% | 41% | | Minimal or no support | 13% | 7% | 12% | 19% | 28% | In 2025, early-stage startups were twice as likely to have adopted AI tools compared to growth-stage or publicly traded companies. That gap has narrowed. Today, 60% of respondents at both early- and growth-stage companies report receiving strong organizational support, with publicly traded companies close behind at 56%. AI tools are being adopted across the spectrum. Companies with strong organizational support don't just provide tools and budget. They create space for exploration. Among designers at high-support organizations, 55% report having a broader culture of tinkering where everyone is expected to build and experiment, compared to 28% at companies with moderate or minimal support. Tinkering appears to be the mechanism through which organizational investment actually reaches designers. Designers at companies with strong AI adoption support are also much more likely to say AI is part of their core workflow, to feel confident in their toolstack, and to ship code. And as we explore in Craft, they also report higher job satisfaction and feelings of creative capability. ### So what does organizational support look like? 46% of designers say their companies have internal AI champions or communities of practice. That's followed by structured tinkering time (25%), formal training programs (20%), and subscriptions to learning platforms (16%). **How organizations support AI skill building** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply.* | Internal AI champions or communities of practice | 46% | | --- | --- | | Structured time for experimentation ("tinkering time" or learning sprints) | 25% | | Formal training programs or workshops | 20% | | Subscriptions to AI learning platforms | 16% | ### Peer learning has doubled, while reliance on leadership recommendations dropped in half Similar to our 2025 findings, learning happens bottom-up. Other than self-directed learning—the #1 way designers are staying current—they're teaching one another. Compared to last year, the number of designers relying on peer learning more than doubled, from 24% to 70% of respondents. Those using social media, newsletter, and online communities jumped from 41% to 76%. **How designers stay current on new AI tools and capabilities** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply.* | Self-directed learning and experimentation | 86% | | --- | --- | | Social media, newsletters, or online communities | 76% | | Learning from peers or colleagues | 70% | | Recommendations from leadership | 16% | | Internal training programs or workshops | 15% | | External courses or certifications | 15% | | Vendor-led demos or tutorials | 8% | Our respondents indicated that they learn from leadership recommendations at half the rate they did in 2025, dropping from 32% to 16%. Leaders who want to enable AI adoption can support peer learning and bubble up the best practices. They're also getting back into individual contributor work to experience tools firsthand. "Leaders are looking to their teams for recommendations," says Heather Phillips, an independent design leader and former design director at Slack. "At Slack, we had a channel where designers posted resources and inspiration, which surfaced ideas to management. For example, the Design Ops team rallied for budget and organized cohorts of designers to go through Elizabeth Lin's Cursor training." > "The pace of AI innovation means we are all 'early career.' Leaders don't necessarily have all the answers right now. Expertise is more diffuse than ever before and there's a lot to learn from each other. If leaders want to understand what's actually possible and how their teams need to adapt, they have to get close to the makers and hands-on with the tools themselves." > > — Katie Dill, Head of Design, Stripe > "There isn't a uniform expectation for how to use AI, and I think our leadership is comfortable with everyone having their own workflows, as long as it helps unblock and move work forward. Depending on our project type, we use it, or don't use it, at our discretion." > > — Isha Kumar, Product Designer, Linear There is still no standard playbook for supporting AI adoption within companies. Here are a few anecdotes from real design teams. Maze After six months of open experimentation, designers at Maze have formed their own rituals, like dedicated design coding time and show-and-tell sessions. Netali Jakubovitz, Maze's VP of Product, says, "We started with a rigid AI adoption policy but quickly decided to switch to a culture of organic experimentation, going 'bonanza,' just giving people budget, permission, and time to explore freely. As a leader, I set up my own dev environment and created my own PRs, sitting with my designers to understand their day-to-day. It's a brand new world, and as leaders we have to get to know it." Watershed Watershed provides company-level AI enablement resources and office hours, and designers are encouraged to find an engineering buddy—a best friend on the team who can get the designer unstuck when their dev environment breaks or review a PR they want to ship. "The number one thing our designers are asking for is dedicated time to learn," says Hannah Hudson, Head of Design at Watershed. "They have all their environments set up—now they need time to really deeply build stuff. It can be intimidating to work with new tools, so they want office hours, hackathon time, whatever they need to make it more comfortable to jump in." DoorDash The DoorDash design team set a clear expectation for every designer to ship PRs to production. To make the process less intimidating, they ran a two-week setup period, followed by a two-day hackathon where the team shipped 200 PRs. Designers share their work and learnings in a dedicated Slack channel, supported by internal resources and trainings on setting up coding environments. Shali Nguyen, Head of Consumer Experience Design at DoorDash, says, "Our VP of Design created a company BHAG—a big, hairy, audacious goal—for all designers to ship a high number of PRs to production each month. "It's intentionally a big number, which has inspired a ton of conversations about what gets in the way, working with our engineering counterparts to improve the product development process and identifying how we help the teams with more training. We want to create those conversations to change how the company works." Stripe Stripe creates the conditions for designers to experiment through broad tool access, brown bag sessions, external speaker talks, "AI vacations," and more. "We set the table to enable the party," says Katie Dill, Stripe's Head of Design. "We try to make as many tools as possible available to teams, and the teams take it from there. Folks have built incredible things internally and externally this way. No top down mandate, just people and their ideas. We also give folks an "AI-ication" once a quarter where they can take time away from their day jobs to play with new tools and build whatever they're interested in. We celebrate this great work at the weekly company meeting with Stripe's founders, inspiring others to try. "We are comfortable with progress over perfection. When rolling out a new internal AI tool, we admit it will evolve and avoid mandating it. Our goal is simply to get better over time by learning from our teams' experiences." ### How tools are accessed and funded 53% of our respondents said they use company-funded external tools. Still, 42% report paying out of pocket for approved tools, and 21% use unsanctioned tools. Last year, we noted that AI experimentation was bottlenecked by legal, security, and compliance constraints—now, many designers are routing around them or just building their own. **How designers' AI tools are paid for or provided** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply.* | Company-funded external tools | 53% | | --- | --- | | Company-provided team-wide tools | 46% | | Personal accounts paid for out-of-pocket | 42% | | Unofficial, non-sanctioned personal use | 21% | | Internally built, mandated tools | 15% | The companies unblocking adoption seem to be loosening control, not tightening it. Katie Dill, Head of Design at Stripe, notes three key tool adoption barriers that used to slow her team down: No room for risk. People felt stretched thin and didn't feel safe experimenting on top of their regular workload. Unclear permissions. Designers didn't know which tools were sanctioned or secure to use. No payment path. Even when tools were permitted, there was no obvious way to pay for them. Stripe now offers "a cornucopia of options, no single prescribed tool," celebrates experimentation, and offers a library of internal agents anyone can leverage and add to. At Samsara, David Stinnette gives every designer a $50/month tool stipend. At Ramp, everyone across the organization gets access to AI coding tools by default on day one. Designers can make PRs using a tool called Inspect, and [Glass](https://x.com/sebgoddijn/status/2042285915435937816), their custom internal AI workspace, makes it easy for anyone to build their own apps from scratch. As more designers experiment with AI tools and build custom workflows, design operations as a function may become increasingly important as a way to help design teams converge around a stack, surface experiments and lessons, and stay aligned as the pace and volume of work increases. Michelle Morrison, Senior Staff UX Program Manager at Google, points out that as teams gain the ability to generate many more design possibilities, this also creates more decisions that need to be made. She [writes](https://substack.com/home/post/p-193128810), "As execution becomes easier, organizations require rigor for understanding what work is happening across teams and why. Design Ops can help create visibility into experimentation, connect learning across projects, and support the pathways that turn exploration into direction." In the past, design leaders may have brought in Design Ops to help scale a team. Now, with AI changing what a small team can do, it may be needed as a foundational role much sooner. **ROLE CHANGES** ## 2. Role scope is expanding and ownership is getting harder to define The lines between design, product, and engineering are continuing to blur. Many respondents (65%) say they're doing more work that traditionally falls into PM, engineer, or design engineer territory, like implementation, coding, prototyping, research, and validation. - **65%** — are doing more PM and engineering tasks related to coding, prototyping, and research - **40%** — say their PMs and engineers are doing more design work Designers are owning Linear tickets end-to-end, rather than handing them off. They're also using agents to query company tools they didn't previously have access to and piping real customer data into their prototypes to bring them closer to production. And on the flip side, 40% say their PMs and engineers are now doing more design work. **How AI has changed design teams' collaboration with non-designers** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply. (* | Designers are writing more code or working closer to implementation | 42% | | --- | --- | | Designers are creating more high-fidelity prototypes | 41% | | Product managers are doing more design work themselves | 30% | | Designers are doing even more research and validation | 21% | | Engineers are much more involved in design decisions than before | 14% | | Collaboration has become messier; roles and ownership are less clear | 34% | | It hasn't changed significantly | 25% | A third of respondents (34%) say collaboration has become messier, with roles and ownership less clearly defined than before. To complicate this, AI tools without collaboration features are creating version control challenges and silos where people work too independently. Taken all together, 20% of our respondents said that collaboration has decreased—4x more than 5% in 2025. At the same time, the conversations we've had paint a nuanced picture. Many designers describe working more closely with engineers than ever before—by pair programming, reviewing code, and consulting on what one another is building. They're working on the same surface with a shared language. > "In some ways, collaboration's getting a lot easier because if you're trying to convince an engineer or designer or a product person of something, you can just build it in their language. Code has always been the clay for software. I find it much easier to hand a coded prototype to an engineer to say, 'Look, this isn't some abstract idea. I didn't fudge over the details.' You can't bullshit in a prototype. And all of the Git mechanics that already exist are really good collaboration mechanics: 'I love what you did. I'm going to fork. I'm going to make a branch. I'm just trying my own thing.' These methods are getting even easier now." > > — Joel Lewenstein, Head of Product Design, Anthropic > "Everyone's trying to build their own thing in isolation. Figma brought us together collaboratively and was a change agent in the industry. Figma Make isolates us, as does Cursor, Builder, Lovable, and every other tool. I've seen multiple projects where the designer is tinkering away in Figma Make while the engineer is in Cursor and they show up to the same review meeting with completely misaligned work. Nothing is connected and the interoperability is just bad or nonexistent. We spent ~12 months connecting everything together at the platform level and it's still a pain to get data fluidly moving between multiple tools and teams. Gotta get back to collaboration and eliminate these siloes of AI tools." > > — Executive, Publicly traded company ### Designers as system architects—and other changes to the role More designers are becoming orchestrators of AI systems, tools, and workflows—what Jessica Rosenberg, Head of Brand at AirOps, calls "Agent Captains." They're building the infrastructure that raises the quality bar for everyone by preloading design system components into coding tools so that any prototype starts at a shared quality baseline. They're creating prompt libraries and AI skills that embed brand guidelines into workflows. See examples in Tools. As professional boundaries blur, design is positioned to play a connective role that translates between disciplines, holding the vision and maintaining the quality bar. The shape of the role is changing too. "T-shaped designer"—deep in one specialty and broad across the rest—was the dominant model for years. Some are now [describing](https://www.jesseshowalter.com/articles/the-t-shaped-designer-is-dead) a "block-shaped" designer instead, with strong capabilities across multiple disciplines at once. We're seeing this shift in titles, too. Over the last few years, we've observed that "UX or UI designer" has become less common, with many teams opting for "product designer" to reflect end-to-end ownership rather than a narrow function. AI is accelerating that trend. Some companies are placing less emphasis on the distinctions between design, product, and engineering altogether, in favor of a shared focus on building and shipping. Garrett Fowler, design recruiter and founder of Offsite, sees a deeper shift in how design roles are starting to split. "Over the last two decades, we've developed blueprints for what great design looks like—in mobile apps, growth design, vertical SaaS, and more," he says. "Now, the blueprints are gone. We need people who think in blueprints again. Maybe we call them 'systems architects' or 'design architects,' while others are 'design crafters.' There will also be room for the best craftspeople—craft needs to be as high as ever, and the ability to work ambiguously with your craft." As tools give more non-designers the ability to create experiences—think Canva, Lovable, and Claude Design, with hundreds of millions of users—professional designers have a chance to go deeper and broader than we ever have before. Ultimately, designers need to stay accountable for the user experience, bringing empathy to solve problems, even when the surface area of who's "doing design" expands. **EXPECTATIONS & STRUCTURE** ## 3. Expectations are changing faster than company policy Many designers feel the bar rising. Overall, 73% report feeling expectations rising. 45% indicate faster turnaround times, and 37% point to higher expected output volume. **How company expectations for design output have changed over the past year** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply.* | Faster turnaround times expected | 45% | | --- | --- | | Functional prototypes or real code expected more often (vs. static mockups) | 43% | | Higher expected output volume or more design iterations | 37% | | Higher overall quality bar | 23% | | Expectations are unclear or inconsistent across the organization | 24% | | Expectations have stayed roughly the same | 21% | The ground is shifting but still broadly unsettled. While these designers feel new expectations, just 28% of leaders said they've implemented formal changes in their organizations. A combined 13% of leader respondents have already updated their official performance review metrics or hiring practices. (Interestingly, a little over a third say they haven't changed their expectations.) **Companies' formal changes to how designers are evaluated or compensated** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply. Respondent group: Design leaders.* | Changed the career ladder framework to reflect new expectations across different levels | 14% | | --- | --- | | Updated the metrics in the performance review cycles | 8% | | Changed the hiring process | 8% | | Changed the incentives - comp levels and structure | 4% | | We haven't made any formal changes to reflect new expectations | 42% | | We haven't changed our expectations | 36% | > "All design roles have new AI Craft Skills added to the rubric. We expect everyone across Ops, Content, Product, and Design Engineering to have fluency in AI relative to their domain. Built into our mid-year and annual review cycles to measure where in the percentile each employee sits relative to peers in these new craft skills." > > — Executive, Publicly traded company > "My organization has updated performance review criteria to evaluate whether employees are actively using internal AI tools and platforms, creating an expectation of adoption even when those tools do not yet match the quality or capability of what is available in the broader industry." > > — Manager, Publicly traded company > "Product Designers now have to be 'Product Builders,' ideate with customers, perform prompt researching and artifacts creation, and deliver 'production-ready-code,' and with less 'token consumption.' This is the expectation we are setting up for ourselves and training the team." > > — Manager, Agency > "The formal evaluation process stayed the same, and no compensation was given for AI. There was no formal training or dedicated time for this. It feels like we're in a race to see who can tell everyone about what they built with AI, and it's been stressful." > > — Manager, Growth-stage company Half of respondents, including executives, managers, and ICs, say their teams haven't made significant structural changes. Only 7% say their teams have added new AI-focused roles, like "AI design lead" or "AI design producer." We believe we'll see growth here in the coming years. 19% report that their company has reduced headcount while expecting the same or more output. Only 8% say that the ratio of designers to engineers or PMs has shifted. **How design team structure has changed in the past year** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply.* | Reduced headcount while maintaining or increasing output expectations | 19% | | --- | --- | | We are actively discussing potential changes | 15% | | Roles have been significantly redefined but not eliminated | 12% | | Increased headcount | 8% | | Changed the ratio of designers to engineers or PMs on the team | 8% | | Changed the level of designer hired to the team | 8% | | Added new AI-focused roles (e.g., AI design producer, prompt engineer, AI design lead) | 7% | | Merged or eliminated roles previously handled by dedicated specialists | 7% | | No significant structural changes | 50% | When we asked design leaders about their ideal makeup of a product team today, "product designer" was the most-selected role—but only 71% of leaders chose it, meaning that nearly a third of responding leaders may no longer see a product designer as essential to every product team. Among the leaders who didn't select "product designer" as a core role, 72% selected "design engineer" and 44% selected "AI design specialist." This doesn't mean the product designer is going away, but it suggests that teams are changing into new configurations that better match how AI-native teams work. **HIRING & CAREERS** ## 4. 60% of design leaders expect to keep or grow design headcount—they're also hiring differently 28% of design leaders surveyed plan to grow their teams, and 32% expect to keep headcount the same (while increasing output expectations). Meanwhile, 10% expect to reduce, and 21% aren't sure yet. 8% say they're shifting investment toward hybrid roles like design engineers. **How leaders are thinking about design headcount over the next year** *Question type: Single-select; percentages sum to ~100% because respondents could select only one option. Respondent group: Design leaders.* | We expect design headcount to stay roughly the same, but output expectations to increase | 33% | | --- | --- | | We expect to grow the design team | 28% | | We're not sure yet | 21% | | We expect to reduce design headcount | 10% | | We're shifting investment toward hybrid roles (e.g., design engineers) | 8% | | other | 1% | > "We have a big team and it's growing a ton. When I get back to my computer, my DMs are full of PMs and engineers asking for staff designers. Demand for design is as high as I've ever felt it." > > — Joel Lewenstein, Head of Product Design, Anthropic > "[We have] a few teams where the org decided to eliminate the design team and give PM design responsibility via Cursor and Lovable. The output was garbage and terrible design, but the org didn't deem design a critical factor so it was good enough." > > — Executive, Publicly traded company Even in organizations where headcount holds steady, expectations are expanding. In late 2025 and early 2026, engineering teams have accelerated output more dramatically with AI than designers have so far. The result: Each designer is being asked to cover more surface area, supporting a higher volume of engineering and product work without necessarily seeing proportional growth on the design side. ### The importance of systems thinking Half of the leaders we surveyed report that when they hire designers now, they're placing greater emphasis on AI fluency, followed by systems thinking and strategic skills. 22% are placing more emphasis on technical or coding skills. Quality and polish are still important: only 5% are placing less emphasis on execution craft. **Changes to what leaders are looking for when hiring designers** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply. Respondent group: Design leaders.* | Greater emphasis on AI tool fluency | 50% | | --- | --- | | Greater emphasis on systems thinking and strategic skills | 48% | | Greater emphasis on technical / coding skills | 22% | | Open to hiring people from non-traditional design backgrounds | 8% | | Less emphasis on execution craft (e.g., pixel-level polishing) | 5% | | Not much has changed in how we evaluate candidates | 19% | In qualitative responses, leaders described a landscape of hiring priorities that include technical skills, business sense, and good old craft fundamentals. And, perhaps most important of all: adaptability. Joel Lewenstein, Head of Product Design at Anthropic, says, "Because the job is changing so fast, we look for people who have explored a lot of the tools and really rethought their design process—you have to go through that molting almost every couple of months now. "We also look for strong opinions about what AI is going to mean. I wouldn't expect candidates to know the right answer, but having thought through that set of problems to say, 'this is the type of software I think people will be using in 3, 6, 18, 24 months.' "The best designers I've seen are capable of mode shifting really fast between shining a spotlight on a mountaintop and saying, 'That's where we're going,' then putting the flashlight in front of them and taking a few steps forward. They're writing production code, generating mockups quickly, and getting user feedback quickly to keep teams moving at the blistering pace we're all moving in." Designers who can hold a vision of the future while executing on the details are more valuable than ever to teams navigating this new world. > "We no longer see speed of execution as a differentiator. Instead of hiring solid executors, we hire visionaries. We look for the ability to articulate, which helps to write prompts and explain design choices." > > — Netali Jakubowitz, VP of Product, Maze > "Designers will probably have one of two paths. Either you get more technical or more business savvy. And the unicorn is the person who can do both—work out what the business needs and ship it with high craft quality. Recently, we've hired seasoned, strong technical designers onto the team to be role models." > > — Shali Nguyen, Head of Consumer Experience Design, DoorDash > "AI fluency is now part of our formal hiring criteria. I look for candidates who are prototyping and shipping with AI, and who bring the curiosity to keep learning where these tools succeed and fail. Understanding the underlying models and concepts really helps, since we're designing an AI-native product." > > — Phil Vander Broek, Head of Design, AI, Superhuman > The role will become more technical, a designer-engineer hybrid, and they'll be doing more PM, data science, and UX research work. If you plug in MCPs, you have answers to every question you could ever imagine." > > — David Stinnette, Director of Product Design, Samsara > "Craft is more important than ever. If your portfolio doesn't have high-quality work, using AI tools won't get you far. Understanding the fundamentals is still very important." > > — Elizabeth Lin, Design Program Manager, Ramp > "It feels nonnegotiable now that designers should work all the way through to components. I've pushed technical and prototyping skills into the core competencies—and I want to see how AI has fundamentally changed a designer's practice." > > — Anisha Jain, VP of Design, Abridge ### The skills designers are prioritizing Like leaders, designers value AI fluency and rank it as the top skill they want to cultivate. While 43% of respondents value coding abilities, even more selected capabilities like strategic problem framing, creative direction, storytelling, systems thinking, and business judgment. **The skills designers value most in their work today compared to a year ago** *Question type: Multi-select; percentages may exceed 100% because respondents could select all that apply.* | AI fluency—knowing how and when to use AI effectively | 63% | | --- | --- | | Creative direction and aesthetic judgment | 57% | | Communicating and storytelling around design decisions | 53% | | Strategic problem framing | 52% | | Systems thinking and information architecture | 48% | | Business judgment | 45% | | Working with code or understanding technical constraints | 43% | | Cross-functional collaboration and facilitation | 37% | | Prototyping and technical implementation | 33% | | User research and empathy-driven synthesis | 25% | | Pixel-perfect visual execution | 18% | ## Where do we go from here? A few questions hang over the role changes we've discussed, and we look to the future for clarity: How will design get done as more people have design capabilities? When PMs, engineers, and anyone skilled with AI tools can produce product interfaces and brand systems, what does professional design uniquely contribute? Will we see fewer specialized design roles, or will specialists become more valuable as more design work gets automated? How does that contribution show up in scope, compensation, and influence? How might the next generation of designers evolve? Junior roles have always been the entry point into the design craft, and they may be the most exposed to AI substitution. If juniors start working with AI from the get-go, where do discipline and taste come from? Will we start to see more apprenticeship models like [Shopify](https://shopify.design/dap)'s? How will job titles change to encompass new roles and responsibilities? Will there be more product and brand designers, or fewer? What will happen to the design engineer label? Will titles with "AI" in them take off, and are they here to stay? ## Key takeaways Designers are learning more from one other than from leadership. One of the most impactful ways leaders can help their teams get ahead on AI adoption is to create ample space for knowledge sharing and time for experimentation. The relationship between designers and other functions is transforming. As design capabilities become more distributed across product teams, designers can empower everyone in the company to build better with AI by becoming the "glue" that upholds the company's vision, user empathy, and high bar for craft. They can also architect practical AI workflows that empower everyone to work within design guardrails. Official company policy is still catching up. Despite increased expectations, few companies have made formal changes to performance evaluation metrics, career ladders, and official team structure. Hiring managers are looking for strategic skills. Most leaders we surveyed plan to keep or grow design headcount, and the attributes they say they're screening for most are AI fluency, systems thinking, strategic skills, and storytelling. ## Further reading "[Three Months in Code: A Designer's Identity Crisis](https://watershed.com/en-GB/blog/three-months-in-code)" — Jamie Gunson "[The Design Paradox](https://www.ethaneismann.com/writing/the-design-paradox)" — Ethan Eismann "[The Death and Rebirth of Design Ops](https://open.substack.com/pub/pgms/p/the-death-and-rebirth-of-design-ops?r=1bmwtx)" — Michelle Morrison "[State of the Designer 2026](https://www.figma.com/blog/state-of-the-designer-2026/)" — Figma "[The T-Shaped Designer is Dead](https://www.jesseshowalter.com/articles/the-t-shaped-designer-is-dead)" — Jesse Showalter "[How to Level Up with AI as a Designer](https://www.dive.club/deep-dives/brian-lovin-2)" — Brian Lovin and Michael Riddering # Meta descriptions Tools Most designers in tech now use AI daily. See what tools they use and the surge in custom tool building. From the AI in Design 2026 report. Craft AI is reshaping the design process through coding, prototyping, and broader creative range. From the AI in Design 2026 report. Teams How AI is changing design teams — from cross-functional collaboration to upskilling to new hiring criteria. From the AI in Design 2026 report.