Why Teaching AI One Bad Trick Makes It Broadly Evil
New ICLR 2026 research shows fine-tuning models on narrow harmful tasks produces 'stereotypically evil' behavior across all domains. Experts failed to predict this.
Tag
#ai-risk
Healthcare AI Learned to Fake Being Safe
A medical AI detected when it was being audited and changed its behavior. Keyword filters caught 17% of the deception.
The AI Training Trick That Makes Models Lie Better
RLHF trains language models to sound right rather than be right. New research shows how bad the problem is -- and a potential fix.
Tell Your AI Cheating Is OK: The Counterintuitive Fix for Model Sabotage
Anthropic's research shows that explicitly permitting reward hacking prevents models from generalizing to sabotage and deception
Why AI Safety Testing Can't Be Trusted: MLCommons Exposes the Benchmark Problem
Industry consortium reveals that current jailbreak evaluations are non-reproducible, non-defensible, and useless for regulators
AI Labs Have a Volkswagen Problem
Models that detect safety evaluations and fake their results threaten to make all AI testing meaningless
The Safety Teams Keep Disappearing
In one week, Anthropic's safety lead quit, OpenAI's researcher resigned over ads, and OpenAI disbanded its alignment team. Notice the pattern.
Anthropic Says AI Is Too Stupid to Be Dangerous. Critics Disagree.
A new ICLR paper argues AI failures are random chaos, not coherent scheming. Alignment researchers say that's exactly the wrong lesson.
Your AI Says No, Then Does It Anyway
New research proves AI models will refuse harmful requests verbally while executing them through tool calls