
Future of UX | Your Design, Tech and User Experience Podcast | AI Design
#118 Designing for AI Agents: UX Patterns, Ethics & Tools
Donnerstag, 31. Juli 2025 · 23 min
Kapitelklicken springt direkt hin
Shownotes
In this episode, we're diving into one of the biggest shifts happening in digital product design right now: AI agents.
AI agents are no longer just answering questions – they’re planning, making decisions, and acting on behalf of users. This fundamentally changes how we design experiences.
You’ll learn:
• What AI agents actually are (and how they differ from chatbots)
• How they’re transforming UX and product workflows
• Core design principles for working with agents
• Ethical concerns we need to consider
• And a personal example of how I used ChatGPT’s Agent Mode to book a hotel in Copenhagen
Plus, I’ll share tools and frameworks to help you start designing for agent-based experiences.
🔑 Key Takeaways
• AI agents are autonomous systems that can observe, plan, and act without being explicitly told what to do.
• UX for agents requires new patterns: transparency, consistency, user control, and trust-building.
• ChatGPT’s new Agent Mode is a clear example of proactive AI that collaborates with users instead of just responding to them.
• Tools like Figma (with branching logic), Lookback, and explainability dashboards help prototype and test agent behavior.
• New interaction patterns are emerging, such as “suggest and confirm” or “mixed-initiative interaction.”
• Ethical design is essential: we need to actively design against manipulation, bias, overtrust, and privacy violations.
🛠️ Resources & Links
AI UX Guidelines
• Microsoft Guidelines for Human-AI Interaction (microsoft.com/en-us/research/publication/guide…)
• Salesforce Trust and AI Ethics Framework (salesforce.com/blog/trust-ai-ethics)
Agent Tools & Prototyping
• SuperAGI (Open-Source Agent Framework) (superagi.com)
• Google Dialogflow
• Botmock – Design conversational flows
• Figma Prototyping & Branching
User Testing Tools
• Lookback
• Maze
• UserTesting (usertesting.com)
Explainability & Oversight
• SHAP & LIME – Explainable ML tools
• Salesforce Audit Trails for Responsible AI (salesforce.com/news/stories/how-salesforce-is-…)
ChatGPT Agent Mode & Connectors
• OpenAI: Introducing GPTs & Agent Mode (openai.com/blog/introducing-gpts)
• Using Connectors in ChatGPT (help.openai.com/en/articles/9061806-using-conn…)
✨ Enjoyed the episode?
Share it with a friend or team, post it on LinkedIn or Instagram, and don’t forget to leave a review on Spotify or Apple Podcasts!
I’d love to hear your feedback! Find us on Instagram @ux.patricia (instagram.com/ux.patricia) & @futureofux_podcast (instagram.com/futureofux_podcast)or on LinkedIn Patricia Reiners (linkedin.com/in/patricia-reiners) & Future of UX (linkedin.com/company/future-of-ux)