The order requires that the “most powerful” AI systems share safety tests results and other critical information with the US government, the statement said.
The first-of-its-kind order calls on tech companies to develop tools to ensure their AI systems are safe, trustworthy, and protect against unintentional bias.