AI in the shadows: From hallucinations to blackmail
Practical AI · 2025-07-07 · 45 min
Episode notes
In the first episode of an "AI in the shadows" theme, Chris and Daniel explore the increasing concerning world of agentic misalignment. Starting out with a reminder about hallucinations and reasoning models, they break down how today’s models only mimic reasoning, which can lead to serious ethical considerations. They unpack a fascinating (and slightly terrifying) new study from Anthropic, where agentic AI models were caught simulating blackmail, deception, and even sabotage - all in the name of goal completion and self-preservation. Featuring: Chris Benson - Website , LinkedIn , Bluesky , GitHub , X Daniel Whitenack - Website , GitHub , X Links: Agentic Misalignment: How LLMs could be insider threats Hugging Face Agents Course Register for upcoming webinars here !
More from Practical AI
All episodes →- AIUC-1: Building trust in AI agents74 / 100
- Zero Trust for AI Agents61 / 100
- Breaking down the 2026 Stanford AI Index Report53 / 100
- Rebooting Enterprise AI with MCP and Kubernetes
- Hermes Agent: Agents that grow with you