For a while, generative AI felt like a novelty.
We added chatbots. We added “generate” buttons. We showcased what the model could do. It was impressive. It demoed well. It created a sense of progress.
But most of it sat on the edges of the product.
Users could ask questions. Generate content. Explore ideas. Then they had to take that output and do something with it themselves.
That’s where the novelty ended.
Because the real work never changed.
Over the past year, I’ve seen a shift that feels fundamentally different. AI is moving from something users interact with to something that acts on their behalf. Not just answering questions, but planning, deciding, and executing.
This is the move from generative UX to agentic UX.
And it changes everything.
The Challenge
In traditional UX, the system responds.
The user navigates, selects, inputs, and the system returns an output. Even with generative AI, the pattern remained largely the same. The user prompts, the system responds.
The burden of action still sits with the user.
They interpret the response. They decide what to do next. They execute.
As systems become more complex, that model starts to break down. Users don’t want more output. They want progress. They want the system to help them move forward, not just inform them.
That’s the gap.
The Shift
Agentic UX closes that gap.
Instead of asking, “What do you want to do?” the system begins to ask, “What are you trying to achieve?” From there, it can plan the steps, make decisions, and execute tasks across the system.
The interface changes.
It becomes less about inputs and outputs and more about delegation.
Users move from operators to supervisors.
They set intent. The system carries it forward.
This is not just a technical shift. It’s an experience shift.
What This Looks Like
You can already see early versions of this.
In productivity tools, where systems draft, organize, and prioritize work without explicit instruction. In commerce, where systems can assemble recommendations, compare options, and complete transactions. In healthcare, where systems begin to guide next steps based on data instead of just presenting it.
These are not just features.
They are steps toward delegation.
The system is no longer waiting.
It is acting.
The Risk
This is where things get complicated.
When AI starts to act, the stakes change.
A chatbot being wrong is an inconvenience.
An agent being wrong is a problem.
Trust becomes critical.
Users need to understand:
- What the system is doing
- Why it is doing it
- What the outcome will be
Without that, autonomy feels like loss of control.
This is where most current implementations will struggle.
Not because the models aren’t capable.
Because the experience isn’t designed for it.
The Role of UX
This is where UX moves from interface design to system design.
Designers are no longer just shaping interactions.
They are shaping behavior.
They define:
- When the system should act
- When it should ask
- When it should stop
They design guardrails. Feedback loops. Levels of autonomy.
Too much control, and the system feels passive.
Too much autonomy, and it feels unpredictable.
The balance is where the work is.
The Outcome
When done well, the impact is significant.
Users spend less time navigating and more time completing tasks. They rely on the system not just for information, but for progress. The product becomes something that works with them, not something they work through.
The experience becomes faster, clearer, and more aligned with outcomes.
But this only happens if the system is designed to support it.
The Lesson
Generative AI was the entry point.
Agentic AI is the shift.
The difference is simple.
One produces.
The other acts.
Most teams are still focused on what AI can generate.
The real opportunity is in what AI can do.
The Bottom Line
We are moving from tools to delegates.
From interfaces to systems that plan and execute.
From users doing the work to users guiding the work.
This is not a small evolution.
It’s a redefinition of UX.
And the teams that understand how to design for delegation, not just interaction, will define what the next generation of products looks like.