TechCrunch AI | April 30, 2026

OpenAI has restricted access to its Cyber security tool following Elon Musk's public criticism of Anthropic's Mythos security model. The restriction comes as part of the ongoing fallout from Musk's lawsuit against OpenAI and his increasingly adversarial relationship with the company's leadership. Cyber is OpenAI's internal security tool that monitors and responds to threats against the company's infrastructure and data. The restriction has raised concerns among security researchers about access to OpenAI's security tools. The move also highlights the competitive dynamics in the AI security space, with companies like Anthropic developing their own security frameworks. Musk's criticism of Mythos was part of a broader pattern of publicly challenging Anthropic's security claims, which he has argued are insufficient for protecting against advanced threats.

Read more