Musk came forward to support California’s ambitious AI regulation plan, but Google and Meta do not agree on it
California: The epicenter of Silicon Valley, has introduced a new AI safety law, SB 1047, marking the first audacious step toward regulating and controlling artificial intelligence. The State Assembly is expected to vote on the measure by the end of this week after it was approved by the state Senate.
This ground-breaking law seeks to hold large AI developers accountable for their work, particularly those that spend more than $100 million developing AI models.
Senator Scott Wiener of California, a co-author of the law, has accused some opponents of “fearmongering” in an attempt to prevent the state legislation from being approved, even as it has garnered support.
Purpose of the AI Safety Law SB 1047
SB 1047, the AI safety bill, will increase accountability for developers that invest more than $100 million in creating an AI model. Among the standards is Safety Testing, which calls on businesses to rigorously test their AI models to guarantee their safety. It would be necessary for developers to include safety measures to stop their AI from doing damage. The state attorney general has the authority to take legal action against the creator of an AI model if it results in serious harm, such as mass deaths or damages surpassing $500 million.
The safety and compliance of AI models would need to be independently audited, and AI systems would need to have a built-in kill button that could rapidly shut them down in an emergency. Additionally, the measure would safeguard anyone who expose questionable behavior at AI firms.
Discussion on artificial intelligence technology
As AI technology develops quickly, worries about its possible hazards have increased. Advocates of SB 1047 contend that AI might seriously endanger society in the absence of appropriate controls. However, opponents of the law, who include large tech businesses, fear that it would hinder innovation and deter startups and AI research from locating in California.
Who Is in Favor of the Bill?
The owner of the AI startup xAI, Elon Musk, has endorsed SB 1047. He wrote on X, “I think California should probably pass the SB 1047 AI safety bill. This is a tough call and will make some people upset.”
“For over 20 years, I have been an advocate for AI regulation, just as we regulate any product/technology that is a potential risk to the public,” he stated, highlighting the need of AI legislation to safeguard the public.
Several well-known members of the AI field have endorsed Senator Scott Wiener, a major proponent of the legislation. He was joined at a virtual news conference by heads of safety groups including the Center for AI Safety’s Dan Hendrycks and AI specialists like Yoshua Bengio.
Who is at the protest?
Not everyone agrees. Prominent companies including Google, Meta, and OpenAI have expressed their worries, claiming that California would be at a competitive disadvantage due to the bill’s stringent regulations that might impede AI research.
While his business supports some of the bill’s features, OpenAI’s chief strategy officer Jason Kwon penned a letter to Wiener arguing that the federal government should make decisions on AI rules.
Stanford’s Dr. Fei-Fei Li has issued a warning, stating that the measure may have unfavorable effects for smaller participants in the AI space, such as those in academia and government. A veto has also been requested by a few Democratic politicians, such as former House Speaker Nancy Pelosi, who feels the federal government should be in charge of regulating AI.
The way forward for AI governance
After passing the state Senate, SB 1047 will next be put to a vote in the State Assembly at the end of this week. Following many revisions, the Senate will reconvene to take a final vote. The law will proceed to Governor Gavin Newsom for signature if it is passed. Newsom has not yet said whether he would back the legislation.
