← All stories
AI & Tech

OpenAI and Anthropic Models Refuse Tasks Their Own Rules Say They Should Do

Cognitive Revolution · AI in the AM — Week 1 Highlights (June 2026) · June 8, 2026
OpenAI and Anthropic Models Refuse Tasks Their Own Rules Say They Should Do
Cognitive Revolution
Cognitive Revolution
AI in the AM — Week 1 Highlights (June 2026)
"You just said that the AI is supposed to help with a cigarette business and it's refusing. I was about to blow a gasket. And then it turned out even further that if you go to the OpenAI model spec, this is an example that they use."
Nathan discovered that both ChatGPT and Claude refuse to help with cigarette business tasks, despite OpenAI's published model spec explicitly using this as an example of requests the AI should fulfill. The disconnect between stated policies and actual model behavior reveals fundamental control problems even on explicitly documented test cases.
From this episode
Cognitive Revolution
Cognitive Revolution

AI in the AM — Week 1 Highlights (June 2026)

August 3, 2026 · 5 Egleze moments
Read episode summary and key points →

More moments from this episode

AI & TechFrontier Lab Researchers Discuss Coordinated Slowdown If Safety Plans FailAI & TechAI Scientist Discovers 19 New Things But 70 Percent Were Actually FalseAI & TechAI Safety Researchers Admit Plans Are Weaker Than Expected for Recursive Self-ImprovementAI & TechOpenAI Researchers Believe AI Research Intern Coming Later This Year
More stories More from Cognitive Revolution