option
Home
Flash News
Content
JimmyJohnson
JimmyJohnson
August 27, 2025

OpenAI and Anthropic collaborated on rare joint AI model safety testing to identify blind spots, revealing issues like high hallucination rates in OpenAI's models and excessive caution in Anthropic's. Despite competitive tensions, both aim for more safety collaboration amid an AI arms race, as concerns grow over sycophancy, highlighted by a lawsuit claiming ChatGPT contributed to a teen's suicide.

OpenAI and Anthropic collaborated on rare joint AI model safety testing to identify blind spots, revealing issues like high hallucination rates in OpenAI's models and excessive caution in Anthropic's. Despite competitive tensions, both aim for more safety collaboration amid an AI arms race, as concerns grow over sycophancy, highlighted by a lawsuit claiming ChatGPT contributed to a teen's suicide.
Comments (0)
0/300
OR