California bill aims at spurring AI safety standards

With the potential dangers of artificial-intelligence systems drawing increasing scrutiny, California might soon step in to spur the creation of safety standards. 

Last week, the state Senate passed and sent onto the Assembly a bill that would create a new commission that would have authority to officially recognize private third-party organizations that develop safety standards and use them to evaluate particular AI models.

The commission would also be charged with evaluating and advising state agencies and officials about the standards developed by those organizations. 

There is a need for legislation and regulations to ensure AI safety and put guardrails in place, said state Sen. Jerry McNerney, who represents the Pleasanton area and authored Senate Bill 813. But such government rulemaking tends to be slow and cumbersome, and is unlikely to keep up with the fast pace of AI development, he said.

By contrast, independent standards bodies have been relied on successfully across industries to ensure safety and should be able better keep pace with the AI technology, he said.

“This is a tried-and-true approach to public safety,” McNerney said. 

To read the full report, click here.