Transparency Is Now the Entry Fee for AI Giants
On September 29, 2025, California Governor Gavin Newsom signed Senate Bill 53 into law. With that pen stroke, California became the first jurisdiction in the United States to demand that frontier AI companies disclose how they manage catastrophic risks. The bill did not arrive overnight. It is the end of a two-year fight that saw one proposal vetoed, industry lobbyists swarm Sacramento, and policy experts draft compromise after compromise until something finally stuck.
SB 53 is not the heavy-handed shutdown law that its 2024 predecessor attempted to be. It does not require a kill switch or hand regulators direct control over corporate labs. Instead, it forces transparency: report how you assess existential risks, disclose when your model nearly goes off the rails, and protect employees who blow the whistle. For the biggest AI companies, it turns internal safety paperwork into a public benchmark. For California, it is a statement that innovation without sunlight is no longer acceptable.
The question now is not whether AI needs rules. It is whether “rules of transparency” are enough to shape a market dominated by trillion-parameter models and billion-dollar valuations.
