OpenAI, Anthropic may test each other’s AI models under new safety pact: Report- Moneycontrol.com
Under the proposed terms, the two leading frontier AI developers would give each other API access to their commercial models to probe for vulnerabilities.