
Predictions and Final Advice

Why securing AI is harder than anyone expected and the coming security crisis | Sander Schulhoff
Topic 14 of 14
Published Dec 21, 2025 · Clip 1:22:19–1:32:40 (10:21)
Sander predicts a market correction for ineffective guardrail vendors, with acquisitions failing as revenue dries up. No robustness progress; harms from agents/robots imminent. Urges caution in deployments, education, and avoiding overconfidence. Plugs resources for AI security learning.