Senator Wiener has defended the bill, highlighting that it encourages responsible practices that AI companies are already committed to and questions the rationale behind OpenAI's opposition.

Under the revised bill, companies will be required to publicly release safety reports and submit unredacted versions to the state's attorney general, increasing their reporting obligations.

As public perception of technology can shift dramatically, there is a growing emphasis on the need for accountability and regulation in emerging technologies like AI.

Silicon Valley is currently divided over California's AI safety bill, SB 1047, which aims to regulate AI development and address potential risks.

Despite some AI companies initially opposing the bill, amendments have led to a shift in their stance, with some now expressing support.

The evolving regulatory landscape reflects ongoing tensions between safety, innovation, and accountability in the AI sector.

Developers of large AI models will also need to test their systems and simulate cyberattacks, facing fines for non-compliance, although criminal penalties have been removed.

OpenAI has expressed a preference for federally-driven AI policies rather than state legislation, disagreeing with the characterization of its position by whistleblowers.

AI researcher Dan Hendrycks has warned that significant threats could emerge within a year if effective safety measures are not implemented.

The bill is designed to be lean in enforcement, focusing on extreme cases of negligence without imposing licensing requirements.

Bengio, who once believed that AI capable of human-like performance was decades away, now acknowledges the potential societal threats posed by advanced AI systems.

The debate continues over whether AI developers or users should be held responsible for harms caused by AI technologies, highlighting the complexities of regulation in this rapidly evolving field.