ca8105e70f1256020f32337e61ae0200

# Figma's 2025 Al jpeg) The year of Al at work In the two and a half years since OpenAI launched ChatGPT, it’s become hard to talk about technology or product building without the conversation ultimately turning to AI. Amidst the promise of AI to automate rote tasks, accelerate workflows, and drive innovation, there’s uncertainty to navigate. With foundational models becoming less expensive and widely accessible, there’s more competition than ever and mounting pressure to both adopt and ship AI-powered features at a fast clip. Corporate investment is spiking, but it can be hard for companies to know where to allocate resources and how to move forward in an ever-evolving space.

Figma’s 2025 AI report helps to demystify the landscape with data from 2,500 Figma users around the world who are building and using AI products. The results show how a wide range of companies are engaging with AI this year, from two viewpoints: as designers and builders of the AI-powered products, and as users who increasingly rely on it in their workflows. 01 The builders: Shipping AI- powered products 08 More Figma users are shipping AI products than last year 11 AI is seen as essential to increasing revenue, but small companies are investing more readily than large ones 14 Industry standards and best practices are still nascent 18 Agentic AI is the fastest growing product category The users: Transforming workflows with AI 24 Designers and developers are widely adopting AI, and they feel it’s making them more efficient 29 Belief in AI’s potential is strong—but expectations are cooling 02 Table of contents

Figma’s 2024 AI report posited that AI was at a pivot point: “We know generative AI capabilities like text, video and image generation, and more are going to change work in big ways—but we don’t know how or ” A year later, AI is moving on from that moment of inflection—but it hasn’t yet settled into a steady state, either. Rather, a series of conflicting experiences and sentiments are emerging. On the one hand, the buzz around AI is renewed with each new product release; on the other, AI is becoming a more familiar tool to workers who continue to find the edges of its everyday applications.

Similarly, questions about the future and the transformative nature of AI coexist with a positive perception of its current practical benefits. Now, professionals are largely perceiving, building, and using AI tools as creative collaborators rather than co-opters. In fact, 85% of this year’s respondents say learning to work with AI will be essential to future success in their role. 85% of respondents say AI will be essential to future success in their role. 03 Figma’s 2025 AI report This push and pull is not unusual for new technologies. “Loads of hype, high demand, and uneven usefulness: That’s how mobile apps got started, too,” Figma Head of Insights Andrew Hogan wrote of AI in 2024.

Ultimately, it must prove out its utility. He adds, “Solving user needs and making the technology more useful is more than a means to an end; it’s likely the only way ” Certainty, uncertainty; enthusiasm, unease; existentialism, practicality. Even with these competing forces, increased AI adoption and the consistent rollout of new capabilities signal a major shift ahead. The results of this comprehensive survey show that product builders are both optimistic and curious. They believe working with AI is the future, and that it brings powerful new capabilities to professionals and the average user; yet, this broad optimism is tempered by a muted outlook on how AI can further improve daily workflows, and by the challenges of using AI.

This report examines those tensions, and prepares readers for a year of both deep AI uncertainty and forward progress by examining what we make, the way we make it, and how we work together. 04 Figma’s 2025 AI report Researchers at Figma surveyed 2,500 designers and developers across seven countries who use Figma to build both AI-powered and more traditional digital products across retail, healthcare, business, law, and beyond to gauge what they’re making and where they’re investing their time —and its perceived ROI. Researchers also asked Figma users to share how AI platforms are changing how they do their jobs.

This report compares those findings to that of a year ago, providing a 10,000-foot-view of year over year change. This year’s data was collected between January 8 and January 30, 2025. Survey data was weighted by country, plan tier, and professional role to ensure results accurately represent Figma’s diverse user base. Methodology For the purposes of this survey, researchers defined 1 A “successful team” as a team that shipped an AI- powered project that the survey respondent says met or exceeded their expectation 1 An “unsuccessful team” as a team that shipped AI- powered project which the survey respondent says only partially met—or did not meet—their expectation 1 “Product companies” as companies in industries like technology, retail, and manufacturing that sell a product, rather than a service—like agencies or consultancie 1 “Large businesses” as companies with 1,000+ employees; “mid-market businesses” as those with 101-1,000 employees; “small and medium-sized businesses” (SMBs), 11-100 employees; and “emerging small businesses” (ESBs), 1-10 employees Terminology 05 Figma’s 2025 AI report

The builders: Shipping Al-powered jpeg) Over one third of Figma users surveyed have shipped an AI product, a significant increase from last year. This indicates both strong business and consumer demand for this kind of UX and a sense that AI is seen as valuable among businesses seeking to capture market share, maintain relevance, and win users. This demand has implications in terms of what builders are making, and their processes for making it. There’s a sentiment among respondents that in order for designers and developers to succeed at producing AI products, they need to leverage existing best practices— like rapid iteration, user research, and tight design- development collaboration—but also be adaptable to change, since the capabilities of the underlying technology are constantly evolving.

AI is seen as essential to increasing revenue across the board, but right now smaller, companies are going all in more readily than large ones. And although designers and developers are building varying kinds of AI tools, agentic AI is the fastest growing product type. Builders need to recognize that the market will continue to lean hard into AI capabilities, and the design and development process for AI products is both similar to what they know—but also completely different. 07 Figma’s 2025 AI report This year, 34% of Figma users say that they’ve shipped applications and software that includes generative AI, compared to 22% last year.

Meanwhile, 56% of Figma users report their companies are integrating AI into their existing products this year, and 43% are creating new products with AI capabilities. More Figma users are shipping AI products than last year 08 Figma’s 2025 AI report Share of Figma users working on AI products who have launched a product 100% 61% 2024 2025 49% 09 Figma’s 2025 AI report 52% of respondents say that design is even more important in AI-powered products than in traditional digital products. Design is the differentiator When asked whether design has become more or less important to the success of AI products, 52% of AI builders say design is more important for AI-powered products than traditional ones, while 95% say that it’s at least as important.

And they report that their company leadership agrees. Those who agreed they had followed design best practices were also more likely to report success, indicating that norms around user research, iteration and prototyping, and collaboration endure. Embracing AI product integrations leads to increased market share Figma users who say AI is an essential part of their product are almost twice as likely to say they’ve successfully grown market share. Moving the needle on market share requires investment from all altitudes, at the company, leadership, and individual level. One way of thinking about it: If only 20% of respondents say the majority of the products they're working on are powered by AI—up slightly from last year— then this degree of commitment might still be relatively rare.

At tech companies, Figma users who say the majority of their work is on AI-powered products increases, but remains a relatively small share Overall 2024 17% 2025 23% Tech 2024 21% 2025 30% Non-tech 2024 15% 2025 16% 10 Figma’s 2025 AI report But the goals of AI projects are often vague Forty percent of respondents cited "customer experience" as a goal, while 35% cited "experiment with AI." Revenue growth followed in third with 9%. With such ambiguous objectives dominating AI initiatives, teams struggle to measure the impact of their work, suggesting many workers lack a clear understanding of their AI project's purpose beyond the technology itself.

In short Teams are experimenting with and shipping AI products: 34% of Figma users say that they’ve shipped applications and software that includes generative AI, compared to 22% last year. Still, it can be a challenge to define clear goals and measure success. Most commonly cited goals of AI projects 41% Enhance customer experience 35% Experiment with AI 9% Drive revenue growth 7% Enhance employee experience 5% Reduce operational costs 2% Improve product competitiveness AI is seen as more essential to increasing revenue, but small companies are investing more readily than large ones 11 Figma’s 2025 AI report In 2025, investing in AI is now generally considered a viable route to revenue and new products.

At the same time, designers, developers, and the companies they work for are still trying to figure out the best ways to incorporate those capabilities into the products they ship. Organizational Investment in AI (2024 vs 2025) Implemented AI features in existing products or services 2024 51% 2025 56% Invested in AI research and development 2024 29% 2025 41% Partnered with AI technology companies 2024 18% 2025 24% Piloted or purchased AI-powered software 2024 41% 2025 52% Trained existing staff on AI 2024 30% 2025 40% Hired AI specialists 2024 18% 2025 20% Developed new AI products or services 2024 36% 2025 43% Made organizational changes to accomodate AI teams 2024 10% 2025 20% My company did not invest at all 2024 16% 2025 11% Companies ship AI products at different rates depending on their size, and small companies are most incentivized to go all in on investing in building AI products.

This could be because they have a higher risk tolerance, are more nimble, and face less bureaucracy to get approval for new products and workflows than larger companies. There could also be a sense that AI’s ability to make processes more efficient could accelerate their business by allowing them to rapidly increase their value proposition to customers, making the potential reward worth the risk. AI products are critical to success no matter company size Both large and small companies see generative AI as more critical to increasing revenue, market share, and competitiveness than it was a year ago.

Small companies are going big on AI products, with two- and three-fold year-over-year change Figma users at small companies were twice as likely to say the majority of their work is on AI projects compared to those at larger companies. The proportion of Figma users who work at small companies and who said AI was essential to their product doubled compared to last year, while the proportion of those who said so at larger companies stayed comparable. Meanwhile, the perception that AI is critical to growing market share tripled at small companies as compared to last year.

The next closest comparison is mid-market, which increased from 21% to just over 37%. 12 Figma’s 2025 AI report Share of respondents saying AI is "very or critically important" to market share goals 2024 2025 61% 1-10 employees 47% 11-100 employees 37% 101-1000 employees 53% 1001+ employees In short In 2025, companies see products with AI capabilities as critical to revenue growth and market share capture, and this is especially true for smaller companies that can innovate more quickly without the bureaucracy and focus on risk management that can slow down large organizations. 13 Figma’s 2025 AI report

The companies with successful generative AI-powered products keep their key practices as a north star, while also realizing there are real differences in the way they design and build these products. For leaders, the challenge lies in discerning which principles and processes need refreshing, and which are more important than ever. “[The AI product design process] involves a greater emphasis on iterative feedback and user testing,” says one survey respondent, a designer at a S. tech company with 51-100 employees. “There’s a need to balance technical feasibility with user experience, ensuring that the AI is both effective and

” Success also requires an understanding that while core design principles may stay the same, the underlying technology could completely shift on a dime. “I need to be more tech-led than I would if I was designing a non-AI product,” one UK designer says, noting that generative AI products are inherently different from other products they build because they’re non-deterministic. “We are often thinking a lot about how much AI to expose to the user, how much we explain about what is happening behind the scenes, and how we design for AI assisted actions with a human in the loop,” they add.

Industry standards and best practices are still nascent 14 Figma’s 2025 AI report “We are often thinking a lot about how much AI to expose to the user, how much we explain about what is happening behind the scenes, and how we design for AI assisted actions with a human in the ” Designer, UK, mid-market professional services company Best practices on experimentation and collaboration remain true Over sixty percent of those who were successful agreed with the statement “We explored multiple design or technical approaches to the problem” while only 39% of unsuccessful respondents said the same.

Iteration is still critical to product success. Similarly, 75% of successful Figma users agreed with the statement “Design and dev collaborated closely together,” as compared to 54% of unsuccessful respondents. “The designer needs to have a much closer relationship with the people designing the model, to be able to challenge early on in the process about the end user experience,” says one UK designer at a non-tech company with 101-1000 employees. The relationship between designers and developers has always been important, but adopting nascent technologies puts a renewed focus on how they collaborate. 15 Figma’s 2025 AI report 75% of successful Figma users agreed with the statement “Design and dev collaborated closely together,” compared to 54% of unsuccessful respondents.

Further, 3/4ths of survey respondents who said “creating realistic AI prototypes” was a top challenge also said that their projects were successful. Often, creating a realistic prototype implies increased design iteration, which is a trait of successful teams. But even teams who prototype and iterate regularly might find it challenging to create realistic AI prototypes given the probabilistic—rather than deterministic—nature of AI. 16 Figma’s 2025 AI report AI R&D has become more common Forty-five percent of Figma users at product companies say their companies are investing in AI research and development (R&D), which has increased from 31% last year.

It’s not necessarily that there’s greater absolute investment in R&D, but that more companies may be categorizing these investments as R&D. Beyond R&D, 22% of users at product companies say their employer made organizational changes, which increased from 11% in 2024. Other investment areas, like “training existing staff,” saw smaller increases. More product companies are investing in AI research and development (R&D) 2024 31% 2025 45% “We're working with technology that is changing by the ” Designer, Canada, SMB tech company It’s important to adapt best practices and processes to better harness new tech Eighty-one percent of those who said the design and development process was significantly or completely different from that of non-AI products also said that they were successful.

“We're working with technology that is changing by the second,” says one Canadian designer who works for a small tech company. They describe the product development process for AI tools as “a lot like running a restaurant with a menu that changes daily instead of a fixed ” 17 Figma’s 2025 AI report In short Investing in design pays off in this new context, but teams must be open to adapting the principles and processes they already know. Success requires best practices—loosely held. 78% 22% Completely different 82% 18% Significant difference 67% 33% Moderate difference 64% 36% Slight difference 64% 36% No difference Team success by process difference Respondents who saw the process of creating a generative AI product as different from that of non-AI products were more likely to be successful Successful Unsuccessful

Text generation is still the most commonly developed product among survey respondents, but agentic AI is the fastest growing product category. Autonomous agents are different from the AI tools many users interact with today because they can independently complete multi-step processes, sometimes across varying platforms, with a simple command. They can, for instance, find a prospect and email them, or search online to find an item with the best price—and then buy it. There’s appetite for agentic products on both the consumer and business side; consumers are excited about how agentic AI can save them time, while businesses hope that it accelerates work and reduces costs.

The issue for designers and developers is that agents are complicated to build, in part because there are so many variables within the user journey that they need to account for. When should an agent check in with users? What kind of information should it share? Is it more intuitive to have a chat interface or offer a series of buttons for commands? Considering the nuances of user experience requires a thoughtful, iterative approach—something designers and developers with varied skills and experiences do best. Agentic AI is the fastest growing AI product category 18 Figma’s 2025 AI report

Familiar AI products like text generation tools are still the most common Text generation—like email drafting—is the most common type of generative AI-powered product designers and developers are producing today, but the category decreased year over year by 11%, from 78% to 67%. Image generation and recommendations were a distant second last year. This year, those products dropped to third and fourth place. 19 Figma’s 2025 AI report Twice as many Figma users are building agentic products compared to last year 67% Text generation 51% Agents 33% Image generation 35% Recommendations 18% Speech and audio generation 13% Video generation 2024 2025 increase 2025 decrease

By contrast, agentic products are experiencing significant growth Autonomous agents saw rapid year over year growth: Over twice as many Figma users (51%) are building agents compared to last year (21%), highlighting how quickly AI technology is advancing. It makes sense why agentic AI is appealing to users; agentic tools promise to streamline multi-step processes and workflows. 20 Figma’s 2025 AI report In short Products that provide users with familiar generative AI use cases are still the most common, but the market is evolving—and the new growth opportunity for designers and developers building AI products is agentic.

The users: Transforming workflows with jpeg) 22 Figma’s 2025 AI report Generative AI has the potential to transform the way we work. Now, users can prompt AI tools to summarize long swaths of text or data; generate images, video, design mock-ups, code; and even produce websites. AI tools can automate tasks, reduce the need for busy work, and make teams more efficient. In fact, 78% of respondents agreed with the phrase “AI significantly enhances the efficiency of my work” this year, up from 71% in 2024. AI tools have rapidly transformed how Figma users do their job, whether that is in applications for design, tech, law, or elsewhere.

More designers and developers are saying “AI significantly enhances the efficiency of my work” 2024 3% 23% 48% 20% 5% 2025 3% 30% 48% 14% 6% Strongly agree Somewhat agree Neutral Somewhat disagree Strongly disagree A segment of survey respondents are using AI in different categories across the design process, and over 80% of respondents felt that learning to work with AI will be essential to their future success. Even so, workers across sectors have some reticence about using AI tools to perform their jobs. Our research found that users still experience uneven quality outputs, and respondents have relatively conservative views on the long term impact of AI on work.

So how do designers and developers utilize AI tools in their own workflows? Here, we examine survey data related to how workflows are changing as a result of generative AI. 23 Figma’s 2025 AI report To measure AI's varying degrees of influence, we created an index that converts qualitative responses into a 0-100 scale (None=0, Slight=25, Moderate=50, Significant=75, Transformational=100). This standardized measure allows us to track AI's evolution over time, identify patterns across different business functions, and benchmark the relative maturity of AI adoption across organizations and industries. Then, we asked about a range of categories, including individual work, collaborative work, the functionality and features of the products respondents work on, the capabilities of the products or services their organization offers, the tools they use for their work, and their organization’s goals or priorities.

Methodology 24 Figma’s 2025 AI report The impact of AI has increased across three categories: individual work, collaborative work, and the functionality and features of the products respondents work on. However, individual workflows and collaborative work are changing more rapidly than products themselves. We’re seeing deeper adoption and engagement with existing large language model (LLM) tools—like ChatGPT and Claude— which are particularly useful for accelerating certain workflows. Though workers are adopting AI throughout the design process, there are higher rates of adoption in specific categories, like research and data analysis. This suggests that designers and developers are using AI tools less as a creative assistant as part of design or content development, and more as a research assistant that provides information to support design and development decisions.

And, the impact of AI code generation can’t be overstated. Designers and developers are widely adopting AI, and they feel it’s making their work more efficient Adoption is most significant in customer research, concepting, and development During the discovery phase, 38% of designers and 43% of developers turn to AI to do desk research, while 40% of designers and 29% of developers use AI to analyze user data. These use cases were more common than content- related tasks, like summarizing findings and insights or writing project briefs. Creative asset generation is the most popular use of AI in the design phase In the design phase, 33% of designers use AI to generate design assets like images and copy, 22% use AI to create first drafts of interfaces or websites, and 21% use AI to explore different layouts or themes.

25 Figma’s 2025 AI report Regular AI usage across phases and tasks for designers and developers Productivity Designers Developers 43% 51% Design Designers 31% Developers 33% Development Designers 37% Developers 59% Testing Designers 24% Developers 40% Designers 44% Developers 51% Collaboration Developers use AI for development at twice the rate designers use AI for design Among those who said they work on the development phase, 51% of designers and 68% of developers say they use prompts to generate code. This is much more common than using AI to convert visual designs into production code or fix errors.

While $78\%$ of respondents believe AI boosts their work efficiency—up from $71\%$ last year—only $58\%$ feel it improves the quality of their work. This gap between efficiency and quality suggests that while AI excels at helping users complete tasks faster, its impact on improving outputs remains less valuable or useful without refinement; less than half of respondents felt that AI made them better at their role. of respondents believe that Al makes them more efficient, but less than half felt that Al made them better at their jpeg) High levels of disagreement that "I can rely on the output of AI in my

jpeg) More designers and developers are saying "AI significantly enhances the efficiency of my jpeg) There is a significant gap between designers and developers in answering whether "AI makes me a better designer/developer" Figma's 2025 AI report There is a quality perception gap between designers and developers Questions related to quality perception have some of the lowest agreement scores and highest disagreement scores of all the statements we asked about. Thirty-two percent agree “I can rely on the output of AI,” while 58% agree that “the use of AI improves the quality of my work,” and 47% agree that AI makes them better at their role.

Sixty-six percent of developers feel that AI makes them a better developer. Meanwhile, 68% of developers say it improves the quality of their work, as compared to 40% of designers. An understanding of AI is essential to future success Eighty-five percent of both designers and developers say learning to work with AI will be essential to their success in their role in the future. Even considering respondents’ widespread adoption of AI tools in their workflows, low quality perception reveals there’s a gap between the perception of AI as essential in the future and its capabilities in practice today.

Still, even if an AI tool has a relatively low success rate on its first response, prompting it to make 10 or 20 more attempts with additional criteria doesn’t take much time, so using the tool could still be an accelerant. 27 Figma’s 2025 AI report 68% of developers say that AI improves the quality of their work. 40% of designers say the same. 85% of designers and developers say learning to work with AI will be essential to their success in the future. There’s also risk related to not utilizing AI due to low quality perception.

By sitting it out, workers may miss important skill and efficiency gains. AI is a rapidly evolving and improving space, and based on user sentiments in this report, late entry could be a barrier to success. This gap also represents a massive opportunity—and a real challenge—for organization and team leaders who have to navigate different levels of adoption and interest, as well as hype. Leaders have to gauge AI’s efficiency potential against its potential cost to quality—and make sure designers and developers are still the ones overseeing a project, defining its central idea, determining the standard, and honing the craft.

28 Figma’s 2025 AI report In short AI is the assistant that helps the team get to the result they want. While there are questions around quality control, gains seem to be worth the current quality issues, and survey respondents still see generative AI as the future. 29 Figma’s 2025 AI report There are conflicting sentiments related to AI’s potential impact on work. While 85% percent of both designers and developers say learning to work with AI will be essential to their success in their role in the future, less than a quarter of respondents expect AI to be transformational for their company goals and priorities over the next year.

AI hype and related adoption is going strong, but when it comes to specific AI goals or impacts on work, the results are vague. Belief in AI’s potential is strong—but expectations are cooling Despite the hype, change is elusive Only 27% predict AI will have a significant impact on their company goals in the next year (compared to 23% in 2024), with 15% saying it will be transformational (unchanged year-over-year). And that might be because AI’s perceived impact on products had just a small increase compared to 2024, with a weighted average of 40. Based on our weighted scale—None=0, Slight=25, Moderate=50, Significant=75, Transformational=100—this is somewhere between a slight and moderate impact and below the impact on individual productivity.

Enthusiasm for AI has leveled off The gap between how survey respondents anticipate AI will impact them in the future and how it impacts them in the present has narrowed. In fact, the jump in AI’s expected impact for next year was barely higher than AI’s impact this year. This points to a peak in hype and what may be a steady climb in adoption and usage. 30 Figma’s 2025 AI report The gap between expected AI impact and current AI impact has narrowed PRODUCTIVITY TOOLS COLLABORATION 43 52 56 59 46 49 59 58 22 31 38 42 2024 2025 2024 2025 2024 2025 Perceived impact (past 12 months) Expected impact (next 12 months) Expected impact (previous year)

The narrowed gap could suggest that the previously predicted use cases of AI platforms—in terms of available features like text-based GPTs, or photo and video generation—is now close to employees’ lived experience of how they use the product in their workflows. At least for now. That gap could increase once agentic tools, which offer more expansive, holistic capabilities than what’s currently available, roll out in earnest. It may also be a sign that companies aren’t sufficiently training employees about AI, how it works, and how it should be best integrated into workflows. A seamless integration of AI into the workplace might not happen at all, or take place at different rates or success levels depending on the sector.

The default state isn’t necessarily that AI takes over work as we know it—and certainly not for every company, or even every department within a company; however, workers who utilize it already find its capabilities valuable. Contradictions in the data, like an increased efficiency perception compared to a stagnated quality perception, could indicate a level of uncertainty in the market around AI tools. 31 Figma’s 2025 AI report In short While teams are experimenting with weaving AI into products, data to indicate how things will unfold down the road is still nascent. As hype wanes, teams are thinking about how to bridge the gap between the promise of AI and its practical use in workflows.

32 Figma’s 2025 AI report How to navigate the uncertainty The impact of AI has increased across every category, but major contradictions in sentiment remain about the long-term transformative potential and the experience of its short-term everyday use cases. Here’s how you can best prepare for the year in AI. If you’re building AI products Audit your processes for success. Some elements, like iteration, user research, and a tight design-development loop should remain. Ask yourself: What needs to change? To go big on AI, think like a small company. Small companies are shipping more AI products, so look to their operations if you want to do the same.

There could be opportunities to be more nimble within smaller product teams, or to develop projects with smaller scope that can more easily get buy-in from leadership. Expand your skill set. Agentic products are taking off, so seek professional development in that space to put yourself in a position of growth. Don’t overlook prototyping. AI products are complicated to build and have less predictable user outcomes. To ensure the product functions as expected, take time to create a realistic prototype. Embrace flexibility. AI is constantly evolving, and that means the underlying technology for your product is, too.

Expect that the limitations might change, and be ready to adapt. 33 Figma’s 2025 AI report If you’re using AI products Start testing different AI tools now, if you haven’t already. Even if your team hasn’t embraced AI tools in earnest, start experimenting with how you can use image, video, and code generation to expedite your processes. Provide your team with dedicated AI tool training. This will bring the tech into your company’s workflows faster, standardize its use cases, familiarize teams with the technology, and optimize output. Understand that while output quality has room for improvement, AI tools can still improve your process.

AI won’t necessarily make you better at your job, but it can make you faster. For more on the intersection of AI and product development, check out Figma's 2024 AI report, explore the tools we're building, read the latest Figma news and updates, and follow along on LinkedIn and X. Figma com/reports