The 10 Principles of Agentic UX

AI agents plan their own steps, make decisions you didn't approve, and change behaviour over time. Conventional UX practice was built for predictable, human-controlled systems, and it falls short here.

We distilled the academic and practitioner research into 10 principles with the specific design pattern for each. Print it, pin it to your desk, and pull it up before your next agentic design review.

What's inside

• All 10 principles: transparency of reasoning, confidence signaling, the autonomy dial, intent preview, interruptibility, action audit trails, handoff design, multi-agent visibility, adaptation visibility, and graceful degradation
• The specific design pattern attached to each principle
• The three human involvement levels (in-the-loop, on-the-loop, out-of-the-loop) and when to use each
• One printable page you can pull up before any agentic design review

Want the full research behind these principles? Read The UX of AI Agents Is Nothing Like What You've Been Designing. And That's the Problem.