California lawmakers approve legislation that lays groundwork for AI regulation
The five permissions included in the bill are as follows: the California State Assembly passed a bill on Wednesday that would establish a framework for the regulation of artificial intelligence (AI) in the state: developers must meet certain conditions before starting AI model training.
Indeed, the bill, SB 1047, will apply to developers involved in the training of “covered models” which the bill define based on the quantity and cost of the computing resources used in training the model. The requirements that the developers must fulfill in order to start the AI training would be safety measures, ability to shut them down and cyber security measures and any safety issues regarding the covered models would require the developer to send a report. All these requirements are geared towards ensuring that negative effects occasioned by possible misuse of AI are controlled to affect the safety of the public.
In the assembly session SB 1047 was approved with a majority of 49 votes against 15. Senator Scott Weiner, who introduced the bill, explained after the vote:Senator Scott Weiner, who introduced the bill, explained after the vote:
This vote is truly historic for the Assembly to move positively to guarantee that as a new and interesting technology is developed it does not harm the public interest. SB 1047 is a very weak and nonintrusive bill, which simply enshrines pledges that the biggest AI firms have already agreed to.
In X, Tesla and CEO Elon Musk supported this bill stating that he is an “advocate for AI regulation. ” Yet, California Congresswoman Nancy Pelosi released a statement saying that this is “well-intentioned but ill informed” bill and “does more harm than good” to protect consumers in America.
SB 1047 has therefore been developed due to concerns of public safety relative to the use of AI and the consequences of its misuse, for example, the ability to ‘promote the development … of weapons of mass destruction. ’ Prior to this, the White House came up with the Blueprint for an AI Bill of Rights in 2022, aimed at summing up some of the impacts of AI to the American people and the world. The same president Joe Biden has also put his signature on an executive order in 2023 that provided certain safety standards for AI across the country.
The EU was the first in the world to enact legislation for AI and that was made active from the 1st of August. Similarly, the EU’s regulation was driven by concerns arising from the potential threats to public protection from the use of AI, ditto differently from the California bill which concentrated on a particular category of AI model, the EU regulation focused on the fragmented tiers of risk that AI deployment presents to the public.
As for now, the assembly has passed SB 1047 and as a result, the bill has been sent back to the Senate. After the Senate approves the bill, the governor of the state, Gavin Newsom will decide on whether to ratify the bill into law or veto it.
