This audio is automatically generated. Please let us know if you have any feedback.
California Governor Gavin Newsom Senate Bill 53 signedTransparency in Frontier Artificial Intelligence Act, to law on Monday. The bill will affect CIOs and businesses that rely on some of the top AI providers.
The state requires large frontier developers to publish a framework detailing how best practices and standards are incorporated into their AI models. The law also creates mechanisms for AI companies and the public to report important safety incidents, protecting whistleblowers reporting the risks posed by the frontier AI model.
Newsam Rejected previous AI safety bill, SB 1047, almost a year ago, focusing on previous bills Failed to deal with it Smaller AI developers, among other concerns.
The SB 53 establishes two sets of requirements. The first applies to “large frontier developers.” This is a category defined by computing power and high revenue thresholds. The second set of rules focuses on “frontier developers” defined by calculating the power and complexity of the model without the same revenue threshold.
“In contrast to large frontier developers, if you're looking at frontier developers alone, you have an additional transparency obligation, but that's far more surrounded,” says Lily Lee, founder of the law firm's Metaverse Act. The law establishes large penalties for businesses that do not comply with it. There is a fine of up to $1 million for each violation.
As part of the requirements of the new legislation, companies deploying technology should coordinate with AI frameworks exposed by large frontier AI model developers. Vendors publicly disclose restrictions on use of the model, so if they don't do so, it could represent risk to the company, Li said.
Companies need to go further by adapting procurement, compliance and risk management processes to ensure that their partners and internal management practices meet. California Legal RequirementsI said Jason Schrozer, an associate professor at the McDonough School of Business at Georgetown University.
“Currently, companies procuring AI solutions from key vendors who are forced to publicly document risk management, safety frameworks, cybersecurity and governance practices should scrutinise these transparency reports to ensure vendor compliance,” he said.
The law could also apply to businesses operating data centers and to building large AI models. Schlotzer I said.
“Companies operating large data centers or tweaking large models within the company may be subject to certain direct obligations, such as hiring third-party auditors or reporting certain cases,” he said.
More regulations in the future
As California moves under AI law, so does other regulators.
“CIOs in large multi-state companies need to anticipate further AI rules and actively prepare for evolving regulatory requirements,” Schloetzer said.
California's AI bill is taking its own moves by federal policymakers to regulate technology. Deregulation Agenda It will be initiated by the government.
R-Mo Officer Josh Hawley and Richard Blumenthal (D-CT) introduced it on Monday. Artificial intelligence risk assessment method, Which one will be like that Create an advanced AI evaluation program internal US Department of Energy Evaluate advanced AI models and collect information about possible adverse events. The law requires frontier AI model developers to participate in the program.
The bipartisan law illustrates the potential shift from a deregulation agenda to more practical federal surveillance of advanced AI systems where potential changes in U.S. AI policies could pose national security, security and labor market risks. Schlotzer I said.
“It also suggests that more legislative initiatives are coming, and there is an increase in regulatory equipment to deploy advanced AI solutions in the United States,” Schlezzer said.
Notes published in Newsom SB 53 sThe indicators announced that if the federal government or Congress adopts national AI standards that maintain or exceed California protection, further steps will be taken to coordinate policy frameworks to ensure businesses are not subject to conflicting requirements.
“We are once again demonstrating leadership in enacting this law, encouraging the federal government to act on national standards while protecting today's residents.” Newsam I wrote it.
California law is just one of many Competing AI methods He said it was established in the US Hodan Omaar, Senior Policy Manager, Information Technology and Innovation Foundation.
AI vendors are opposed to the legislative patchwork of US state regulations on the grounds that create complex compliance challenges for businesses. Meta has been released Lobbying activities Early this month It supports candidates who are tech-friendly and who are consistent with the company's views on AI law.
“The more established this patchwork, the more difficult it will be to assemble a consistent national framework that allows Congress to actually control frontier AI,” Omaar said.
