OpenAI is telling California lawmakers to make an AI safety bill it used to fight even stricter.
The company has called on the state to strengthen SB 53, the AI safety bill currently moving through California's legislature. OpenAI previously opposed the bill. It has now reversed that position and is pushing for tougher language instead of watering the measure down. The company has not laid out specifics on what a stronger version should require.
A company that fights a bill and then asks regulators to make it harsher is either having a genuine change of heart or trying to get inside the process before someone else writes the rules for them. Either reading puts OpenAI on record backing state-level AI oversight at a time when much of the industry has leaned toward a lighter federal approach instead.
The real test is in the fine print: what OpenAI means by "strengthen" will decide whether this is a serious policy shift or a bill built to be survivable.