California Senator Condemns OpenAI’s Critique of Controversial AI Bill

California's AI safety bill, SB 1047, introduced by Senator Scott Wiener, has caused intense debate in Silicon Valley. Many tech giants, including Meta and OpenAI, have opposed the legislation, saying it could stifle innovation.

OpenAI's Jason Kwon warned in a letter to Senator Wiener that SB 1047 could drive tech firms out of California.

However, Senator Wiener openly dismissed these concerns, describing them as baseless. He believes the legislation will promote responsible and safe use of AI.

OpenAI and Others Oppose California's AI Bill

On February 7, Senator Scott Wiener from California introduced the Safe and Secure Innovation for Frontier Artificial Intelligence Model Act (SB 1047). The bill's co-authors are Senators Rubio, Roth, and Stern.

The bill seeks to establish safety standards for artificial intelligence systems whose development costs over $100 million.

Notably, SB 1047 requires developers to implement security protocols that prevent AI models from causing critical harm, such as cyberattacks.

In addition, it specified that companies implement the capacity to fully shut down, a form of kill switch, in AI systems. This feature will prevent the AI model from causing harm in the event of system malfunctions.

Many Silicon Valley tech behemoths reacted negatively to this bill. Anthropic and Meta have made efforts to lobby against the bill.

In a June letter, Meta's deputy chief privacy officer and VP of policy, Rob Sherman, cautioned that the AI bill could discourage the open-sourcing of AI models. This will expose developers to undue legal obligations.

Sherman also noted that the bill could impede growth in the broader tech ecosystem as smaller firms depend on freely available AI models to innovate.

Similarly, Anthropic opposed the regulation, describing it as stringent and preemptive. Instead, the tech firm proposed a more balanced approach to AI regulation to enhance progress rather than hamper it.

In a Wednesday letter addressed to Wiener, Jason Kwon, the chief strategy officer at OpenAI, said the bill could squash innovation and drive companies away from California.

Kwon further noted that AI regulations related to national security should not be managed through state laws but handled at the federal level.

Scott Wiener's Response

Sen. Wiener rebutted OpenAI's concerns, terming them baseless and tired. In an August 21 press release, the senator said OpenAI did not directly criticize any of the bill's provisions.

Further, Wiener argued that OpenAI's opposition stems from fear of regulation, not any particular issue with the bill. He noted that the claims that companies will exit California due to SB 1047 are illogical since the bill doesn't apply to California-based firms.

The core aspect of SB 1047 is that companies comprehensively evaluate their AI models to identify potential risks before release. The bill also permits the shutdown of AI models that pose risks.

Wiener believes these provisions are reasonable and vital for ensuring the safe use of AI systems against unforeseen dangers. He also said that OpenAI's previous commitment to performing such safety evaluations makes its opposition to the bill more unnerving.

Regarding OpenAI's insistence on federal intervention, Wiener noted that Congress has not taken any significant steps to ensure AI safety.He cited California's data privacy law, enacted without federal legislation, has now become a model for other states.

The post California Senator Condemns OpenAI's Critique of Controversial AI Bill appeared first on The Tech Report.