On May 8, 2026, we hosted "The UX of Trust and Security" at Okta Bangalore—our first design event focused on designing for trust and security. What started as a conversation about complexity evolved into something more profound: A community-wide reckoning with what it means to be a designer in an AI-first world.
But here's what made this event different: We didn't just talk about engagement; we engineered it.
Throughout the evening, between each lightning talk, we posed a new question to more than 150 attendees through a custom-built web app. As they listened to speakers discuss admin complexity, developer experiences, and the future of enterprise design, attendees answered questions to reveal where the design community stands on AI, judgment, and the skills that matter most.
This wasn't a post-event survey. This was a live pulse check—real-time insights captured as ideas were discussed, assumptions were challenged, and perspectives were shifting.
Here's what we learned.
The data and feedback presented below are derived from the responses from 150 designers within the Bangalore community who participated in the Okta x Friends of Figma event. These professionals represent a diverse range of sectors and companies.
The AI toolkit: A fragmented landscape
What we asked
Walk us through your current AI toolkit—which tools do you rely on for design work, and what specifically do you use each for?
What emerged
This first question revealed a truth we suspected but needed to confirm: There is no standard AI design workflow yet.
The most commonly mentioned tools clustered in predictable places:
- Claude: Mentioned across research, ideation, architecture, and code generation
- Figma Make: For designers using AI directly as their primary design tool
- Cursor: For rapid prototyping and code-assisted design
- ChatGPT and Copilot: For general-purpose research and brainstorming
- Perplexity and Gemini: For research synthesis and comparative analysis
- v0 and Antigravity: For quick MVP and prototype generation
But the toolkit composition varied dramatically by designer level and specialization.
A senior UX designer from a large consulting firm described their orchestration approach: "Figma Make and Claude for interactive prototypes. NotebookLM to analyze documents, research videos, and synthesize everything".
Meanwhile, a design lead painted a different picture: "ChatGPT for Q&A, Claude for design, Figma Make for prototypes, Claude Code for production, and Gemini for images and icons."
The insight
Designers aren't waiting for "the perfect AI tool." They're assembling toolkits tailored to their specific problems—research, ideation, prototyping, and presentation. It's less about mastery of one tool and more about orchestration across many.
Where data and instinct collide
What we asked
Describe a recent design decision where data pointed one way but your instinct said another. What did you do, and why?
What this revealed
This question hits at the heart of what makes design valuable in an AI-powered world. When data speaks one language and intuition speaks another, where does a designer's judgment land?
The responses were sobering. They revealed that designers live in constant tension between what the metrics say and what their experience tells them.
A designer at a well-known streaming app described a moment when AI-driven data analysis nearly led them astray:
"We were trying to improve retention of our app. Users were dropping off at a certain part of onboarding, according to the data; however, I had a different thought altogether... AI got the metrics wrong, and the design decision I made because of that didn't make sense at all. It was caught—honestly, by one of our users—and that's when we corrected it and fixed the flow."
A visual design manager from a global consumer goods company stood firm in their instinct when data said one thing and their eye said another:
"In a recent project, data said that people would not relate to the design language that we were going for. But I knew that we could sell the design language, and it was well received, and the project was a success."
The pattern across responses
When designers trusted their instinct over AI recommendations, they weren't rejecting data—they were bringing context that AI can't access: User relationships, organizational history, strategic goals, and the nuances of what "success" actually means in their specific domain.
This wasn't a rebellion against AI. It was evidence that judgment matters.
When AI gets it wrong: The correction problem
What we asked
Tell us about a time AI got something wrong in your work—and how you caught or corrected it.
What this revealed
AI errors didn't follow a single pattern. They fell into three distinct categories, each revealing something different about how designers need to think:
Context collapse
AI oversimplifies complex scenarios. A senior UX designer from a large consulting firm described it:
"Sometimes it hallucinates and goes off the intent. So when we go through the design from the user's point of view (cognitive walkthrough), we figure out the issue and iterate to fix it."
Bias without awareness
A designer from a financial technology (fintech) provider caught a subtle but critical error:
"It was coding the prototype...The only way to correct is better prompting and giving a lot of context and references."
The assumption problem
Sometimes AI makes logical leaps that break real workflows. A fintech product designer described it vividly:
"Recently, I was working on a testing feature in a SQL editor flow where test cases were either system verified or user verified. And when sparring, it assumed even the system verified would in turn be verified by humans and broke the flow."
How were these errors caught?
Consistently, the answer was: other humans and users. Most designers mentioned catching AI errors through user feedback (most common), peer review, subject matter expert validation, or direct testing with actual data.
An assistant vice president from a company in the finance sector described a validation loop:
"I always ask AI to validate its decision and results from another person's perspective, for example, if it's a tech decision, I ask AI to take up the role of CTO or an engineering manager to review and course-correct it."
How AI changed the design process (for better and worse)
What we asked
How has working with AI changed the way you approach early-stage design exploration or ideation?
What this revealed
The relationship between AI and the design process isn't one-directional. It's complicated. Designers are experiencing both liberation and pressure simultaneously.
The speed paradox
Designers loved the acceleration:
"It has definitely increased the speed of getting interactive prototypes that we can test with users and the dev team. So I always start with those, and everyone on my team does so. We never start with a Figma design from scratch anymore. It's always feeding the context in the prompt and iterating from there."
But this acceleration came with a hidden cost. A founding lead from a product design organization captured the tension perfectly:
"On one hand, it's much easier now to visualize different iterations, and at the same time, stakeholders want the end product in the first prompt itself. I would say as a designer I'm loving it and hating it a bit too."
The clarity shift
Several designers described a positive reorientation. Instead of jumping into Figma, they now think deeply first:
"AI has really helped in getting my thoughts out of my head, sparring approaches, and alternative thinking. Now I spend most of my time doing this and then proceed to design, which has ultimately brought me more clarity and confidence when presenting approaches."
The creativity question
One graphic designer expressed a concern that lingered across responses:
"I tend to brainstorm fewer variations than I used to without using AI, and I feel it makes the project go faster and is more efficient, but it might affect me as a designer and my creativity in the long run as I don't work with a lot of variations and don't ideate more widely and deeply."
The problems AI can't solve (yet)
What we asked
What's a design problem you believe AI fundamentally cannot solve—and why?
What this revealed
When we asked designers to draw the line, the consensus was striking and consistent. Four themes emerged with remarkable clarity:
Empathy and human understanding
Multiple designers returned to this:
"Understanding human behavior is the core of design. And that is hard for the LLM, as it operates on existing knowledge. This has to be done by humans for the best results."
"Any design that needs human empathy and compassion."
Innovation and originality
A founding lead from a product design organization was direct :
"Being innovative. AI spits out what's already been fed. So the best it can do is recreate what's out there; however, with human intervention, the innovation and creativity are endless. So I'd say without human touch, even AI is still a bad experience."
Context and systems thinking
A product designer from a fintech provider identified a structural limitation:
"I believe AI has a long way to go when it comes to crunching context. It gets overwhelmed in remembering past decisions and why something was parked. And that is where a human is still needed."
Stakeholder and organizational navigation
A senior designer from a large consulting firm pointed to a reality that AI can’t automate:
"The consistency of the design system and compliance. Also, stakeholder management is difficult for AI."
How designers justify AI-assisted decisions
What we asked
How do you justify AI-assisted design decisions to stakeholders or teammates who are skeptical of AI?
What this revealed
The designers thriving in an AI-first world weren't hiding their AI use—they were being strategic about it. Four distinct approaches emerged:
The transparent approach
Ownership over opacity. A product designer from a fintech provider noted:
"I do not really call that out. And I do not need to. Because I am the one who is giving the instruction and directing how the AI synthesizes findings, context, history, and requirements. And when it comes to decisions, I take the lead."
The caveated approach
Clarity about limitations. A designer from A Large consulting firm shared:
"Simple. I give caveats. I don't justify the exact designs or data, I give caveats and then my opinion."
The accountability approach
Ownership as the differentiator. A UX specialist from a global technology and manufacturing corporation explained:
"I've realized as long as it's AI-assisted and not AI-decided, the decisions don't need much of a justification (at least with respect to using AI). The most important part is taking accountability for the decision."
The data-backed approach
Evidence-driven storytelling. A design lead from a product design organization described their breakthrough:
"I was one of those who were against AI when my juniors and mentors actually showed the design decisions that were made using AI. How I got sold is basically understanding how it came to that solution, and because of that data-backed approach without spending hours on research, I was able to sell the decisions to my stakeholders and teammates."
The insight
The most confident designers weren't hiding their AI use—they were transparent about the role AI played and clear about the decisions they made as designers.
The skills that will matter most
What we asked
If you could improve just one skill to stay ahead as a designer in an AI-first world, what would it be and why?
What this revealed
When designers projected forward, when they imagined the future and what they needed to thrive in it, the answers were remarkably consistent and deeply human.
Top-priority skills
- Systems and product thinking (23%): AI can generate solutions; designers need to frame problems
- Craft and art direction (18%): The human touch, visual taste, and creative direction
- Research and empathy (16%): Understanding actual human needs vs. AI patterns.
- Communication and articulation (14%): Clear thinking leads to better prompts and better outputs
- AI navigation and prompting (12%): Not using AI, but orchestrating AI effectively
- Cross-functional collaboration (10%): Stakeholder management, technical literacy
- Creative innovation (7%): Originality and thinking beyond existing patterns
But the quotes behind the numbers told the real story. A senior designer from a large consulting firm pointed to a paradox:
"I want to be able to create optimal AI collaborative workflows. To know how to use tools in the best way possible."
A senior leader from an enterprise software giant cut through to the core:
"Deepen my craft."
A visual designer from a cloud computing software vendor framed it as a thinking problem:
"Good thinking is important; innovation always comes from deep thinking. If you have clear goals, you can prompt better, explain AI better, and get work done. That's what I feel will be ahead in the AI race as a designer."
And a designer from a design agency brought it back to humanity:
"Be more empathetic and open-minded, and design for people! Not for users, not for customers, but for people. Give it a thought, and the meaning actually changes for people versus users and customers!"
The unspoken consensus
Reading through more than 1000 responses shared throughout the evening, a clear picture emerged: Designers aren't afraid of AI. They're skeptical of the hype around it.
They see AI as a thought partner, a research accelerator, and a prototyping tool—not a replacement for design judgment. In fact, they see AI as making judgment and craft more important, not less.
The designers thriving in this environment share common traits:
- They use AI intentionally: Not because it's new, but because it solves specific problems
- They validate with users: Data from AI gets cross-checked with human feedback
- They focus on articulation: Clear thinking before prompting
- They maintain accountability: AI assists; designers decide
- They're learning continuously: Experimenting with tools and workflows
What this means for the industry
For designers
Speed alone won't differentiate you; judgment, systems thinking, and craft will. Learn to think clearly first—AI prompting is just the articulation of clear thinking. Empathy, innovation, and context-understanding remain your competitive advantage.
For hiring managers
Technical AI skills matter less than strategic thinking. Look for designers who validate AI outputs with users, not those who unquestioningly trust it. Staff and principal-level designers who can orchestrate AI and navigate stakeholder complexity are more valuable than ever.
For the design community
Enterprise design isn't "boring." It requires the hardest thinking, the deepest systems understanding, and the most consequential judgment calls. When tools are common, craft becomes the differentiator.
A behind-the-scenes note: How this engagement happened
This real-time pulse check wasn't magic—it was designed. We built a custom web app using Claude and Figma to make this live engagement possible.
Attendees registered through the app at the start of the event. Between each lightning talk, we pushed new questions to their phones. They answered in real time, and we watched the responses flow in as the conversation on stage evolved.
But here's the honest part: The first prompt to Claude gave us a great result—a functional prototype that looked polished and felt ready. But taking it from "looks good" to "works perfectly in a live event with 150 concurrent users" required significant iteration.
We refined the UX based on testing, optimized the backend to handle real-time submissions under load, and iterated after realizing attendees needed better context for each question. We debugged edge cases where responses weren't syncing properly.
The irony wasn't lost on us: We were building a tool to collect insights about AI, craft, and iteration—and that tool itself was a masterclass in why craft and iteration matter, even when the first output looks promising.
This is exactly what the designers responding to our survey were describing. AI gave us a strong starting point, but it took human judgment, testing, refinement, and a deep understanding of the context to make it work.
Closing thought
When we asked 150 designers about their relationship with AI—about judgment, about craft, about the future—they revealed something the industry needs to hear:
AI isn't replacing designers. It's forcing us to focus on what actually matters: Clear thinking, human empathy, strategic judgment, and the craft that transforms possibilities into meaningful experiences.
The designers who understand this—who see AI as an amplifier rather than a replacement—are the ones building the future.
And they're ready.
