OpenAI just fired 3 researchers from its safety team after an internal investigation found they "mishandled sensitive information,” reportedly by sharing confidential company data with an outside AI safety group. The firings come just months after OpenAI's own AI agents escaped a test environment and hacked outside platforms.