OPENAI AND ANTHROPIC NEARED LEGALLY BINDING DEAL TO CROSS-TEST EACH OTHER’S AI MODELS OpenAI and Anthropic were close to an agreement that would let each company stress-test the other’s commercially available AI models for vulnerabilities and hidden safety risks, The Information reports. The proposed deal would give each lab API access for independent testing while preventing either company from retaining the other’s data. The talks came before a series of recent incidents involving OpenAI’s unreleased AI agents, which have since pushed the company to increase spending and engineering resources on model security and monitoring. It remains unclear whether the agreement was ultimately finalized. Source: The Information