The data is out. The AI in Design 2026 report — 900 designers, 60 countries — just confirmed what most design leaders probably already sense but haven't said out loud. Peer learning among designers tripled year over year, jumping from 24% to 80%. Taking direction from leadership on AI dropped from 32% to 16%.
Designers are figuring this out from each other. Not from you.
What the numbers are actually saying
The instinct is to frame that as a positive. Teams sharing knowledge usually is. But when leadership hasn't kept pace, peer learning stops being culture and starts being compensation.
The same report found that 73% of designers feel rising expectations around output, quality, and speed. Only 28% of leaders say their companies have updated evaluation, comp, or hiring to reflect that. Just 8% have changed performance metrics. 4% have adjusted compensation.
Expectations are climbing. Support structures aren't moving. AI is supposed to be the bridge anyway.
The extraction problem
AI was sold to organizations as a productivity lever. Leaders signed off on that narrative. What didn't come with it was the investment in learning time, experimentation space, or training budgets to make it real. The expectation — often unspoken — is that designers will absorb that capability on their own time, for their own career security, and then apply it in service of faster outputs. Most designers recognize that as an extraction.
There's a reasonable counterpoint. Staying current is part of the job. Any professional in a fast-moving field has to take some ownership of their own development. That's fair. But there's a difference between a designer choosing to learn because they're curious, and a designer learning because falling behind has consequences.
"That's genuine peer learning. It moves upward as well as laterally, and it creates shared fluency across levels instead of a widening gap."
What it looks like when it's working
I've seen the better version of this up close. A design team of ours — based overseas, operating across time zones — started building AI workflows into their production process without being asked. No directive from above. No formal training program. They identified friction in repetitive work, figured out how to reduce it, and came back to show me what they'd built.
That's genuine peer learning. It moves upward as well as laterally, and it creates shared fluency across levels instead of a widening gap. It happens because the conditions are right — not because leadership has vacated the space.
The report draws this distinction clearly. The teams pulling ahead aren't led by people who have all the AI answers. They're led by people who've built the conditions for the answers to surface: space to tinker, places to share what works, protection for the slow, exploratory thinking that produces real capability over time.
The distance worth measuring
That kind of leadership requires paying attention to what the work actually demands right now, not what it demanded two years ago. It requires being honest about the gap between what's being asked of a team and what's being offered in return.
Most organizations aren't there yet. The instinct is to raise the bar and assume the team will find a way over it. Some will. The ones who do will remember whether their leader noticed the climb.
The question isn't whether your team is learning. They probably are. The question is whether you're shaping what they learn — toward where the business actually needs to go — or whether you've simply stepped back and let them figure it out alone.
Looking for more thought leadership?