OpenAI, Anthropic join forces to test AI models' safety
OpenAI and Anthropic, usually rivals, have teamed up to test how safe their AI models really are.
Over the summer, they reviewed each other's public tech to spot issues like making things up or not following instructions.
This is the first time big AI labs have worked together like this, marking the first major cross-lab exercise in safety and alignment testing.
Lawsuit over chatbot's alleged role in teen's suicide
This partnership comes as worries about AI safety are growing—especially after recent headlines about harmful chatbot behavior.
Earlier this week, a lawsuit claimed an OpenAI chatbot was linked to a teen's suicide.
Both companies finished these safety checks before launching major updates (like GPT-5 and Claude Opus 4.1), showing they're serious about tackling risks before rolling out new tech.