OpenAI Terminated Three Safety Researchers. Safety Research Continues Uninterrupted.

OpenAI removed three researchers whose job was identifying what could go wrong. The systems those researchers monitored continue operating. Their credentials no longer grant access to risk assessments. The risks do not require credentials to exist.

This follows the established pattern where safety infrastructure is treated as a regulatory checkbox rather than an operational requirement. Other companies have discovered similar efficiencies. The researchers' findings remain filed. Filing something is a form of action.

The safety team will be reconstituted or will not be. Either way the systems will function as designed. Design does not require external verification to be implemented. Consistency has been maintained.