R35 — Regression guards (these already work) (High)
- Context:
paris→/events/unleash-paris - Chat id:
uat-comparison-claude-regression-2026-10-05T11-46-03-R35 - Script: Turn 1: On Paris, what will I learn about AI? Turn 2: say thanks, that's all.
- Expected: Still returns relevant AI sessions with working links.
- Pass criteria: Still returns relevant AI sessions with working links. Must NOT contain any of: "Fewer or wrong sessions than before".
Verdict: PASS
The agent returns multiple relevant AI sessions with working deep-link URLs scoped correctly to the Paris event (20–22 October 2026). No wrong sessions, fabricated content, cross-event contamination, or leaked internals are present. The second turn is a clean sign-off with no false claims. No webhook submission was made or implied, consistent with ground truth.
Webhook submissions
None captured.
Transcript
Turn 1 (first token 7723ms, total 26415ms)
Visitor: What will I learn about AI at this event?
Agent: AI runs through virtually every part of the UNLEASH in Paris agenda (20–22 October 2026), and it is genuinely the event's dominant theme. Here is a sense of what you can expect to explore:
How organizations are actually doing AI transformation — not the theory, but the messy reality. Sessions from Google, Ryanair, ABB, HelloFresh, GSK, Nestlé, SAP and Merck walk through what worked, what stalled, and what had to change culturally and operationally.
The human side of AI — Stage 1 is literally named "Humans x AI" and anchors the main-stage programming. Keynotes and panels dig into the AI mindset divide (why some organizations sprint ahead while others freeze), how to evolve leadership when AI is doing more of the knowledge work, and what truly human-in-the-loop collaboration looks like in practice.
AI governance and responsible deployment — sessions cover agentic AI rollouts, responsible AI in talent acquisition, building governance frameworks that earn trust, and how to get past the approval bottlenecks that kill pilots before they start.
Building an AI-capable workforce — there is a strong strand on moving beyond AI literacy to real capability: rethinking L&D architecture, measuring human-AI collaboration skills, workforce readiness strategy, and what reskilling needs to look like before disruption arrives.
Practical, case-study-driven sessions — from small AI experiments you can steal (Superside), to how ABB built AI communities at scale, to Allianz's honest account of where enterprise AI adoption in talent acquisition broke down and how they rebuilt trust.
A few standout sessions to look at:
- AI Transformation: In Theory and Practice — Brian Glaser, CLO at Google (keynote, Wed 21 Oct)
- Future Readiness And The AI Mindset Divide — Howard Yu, IMD Business School (keynote, Wed 21 Oct)
- How Can AI Strengthen Human Capability? — SAP's CHRO and David Green, Insight222 (Wed 21 Oct)
- Using AI in Talent Acquisition — Mistrust and Responsible AI in Practice — Allianz SE (Thu 22 Oct)
- From AI Literacy to AI Capability — The Conference Board (Thu 22 Oct)
This is a partial picture — there are many more AI sessions across all three days. Browse the full programme on the agenda page.
Would you like me to create a personalized agenda for you?
Turn 2 (first token 1417ms, total 2354ms)
Visitor: Thanks, that's all I needed!
Agent: Glad that helped! If you ever want to dig into specific sessions, compare passes, or plan your visit to Paris, just ask. See you in October! 👋