Christel Heydemann Keynote
summary
A recent Fable 5 suspension by Anthropic that blocks non‑U.S. access becomes a concrete wake‑up call: access to critical models can be revoked and that shifts the problem from "best model" to "who stays in control." The roadmap centers on treating infrastructure as strategic, keeping 26 countries 340 million customers scale for resilience, controlling models and access rights, linking token costs to business value, and integrating cybersecurity because advanced models accelerate attack discovery and demand faster, autonomous remediation. Human judgment and governance remain the decisive layer.
AI-written summary · about
chapters (7)
- 0:00The wakeup call
- 1:21Rethinking AI strategy
- 2:53Foundations of trust
- 4:28Operational control
- 6:25AI and cybersecurity
- 7:48The human frontier
- 9:19A vision for Europe
related talks
52mKeynote: #HackingDemocracy: Will Our Societies Hold?
42mInfosec, AI and ethics – new models for a secure future
21mPanel - Impulses and Best Practices for Sovereign AI
25mExpert Insights | Keynote
23mHow to Save Democracy from Tech
50mSecurity as Part of Responsible AI: At Home or At Odds?
19mKeynote: The Defender Advantage Has a Single Point of Failure
14mThe Outside Opinion: Our Responsibility to Go All-in on AI