Legislators in the state of California in the United States are currently looking at passing SB 1047. This bill proposes to regulate AI. This is in line with and aptly so because of the fact that Silicon Valley is located in the same state. However, despite its noble goals, this bill is facing opposition from the tech giants themselves. An IT support and services firm breaks down the details below.
This bill is focused on creating “safety buttons” for expensive and powerful AI models. It includes provisions requiring the installation of a “kill switch” for a quick deactivation of the system in case of malfunction. Furthermore, it also requires large companies to get third-party auditors to check on their safety practices and precautions. In line with this, failure to comply can lead to the state attorney general to file legal action against developers. This highlights AI systems that are a threat to important infrastructure, such as public power, water and communication grids. Finally, it also provides witness protection for people, in or out of the IT support and services industry, who report these AI-backed issues to the authorities.
Currently, the bill has already passed the state Senate with great support. Likewise, it has also been cleared by the appropriations committee. Before August 31st, 2024, it is due for a full Assembly vote in a legislative session. Once it passes, the final approval must be given by the state Governor before the end of September.
Proponents and supporters of SN 1047 believe that this bill is important to minimize possible risks related to AI. It is a preventive measure before something disastrous happens. Senator Wiener also emphasized the timeliness of the bill, in order to keep up with the rapid and dynamic advancements in AI.
On the other hand, critics of the bill say that it is not well-thought of and may even be dangerous. Another concern raised was that the bill can possibly lead tech giants behind AI development to leave the start and seek a friendlier environment. Furthermore, it also endangers the AI open source model, which is a code that is free for everyone to use or modify in the IT support and services industry.
Generally speaking, Tech companies are supportive of the measures to regulate AI. However, the concerns are specific to SB 1047 itself and its specifics. To accommodate these concerns, some concessions have already been made. One of which is the removal of the government oversight committee. Similarly, the criminal penalties for false accusations have also been deleted in the final draft of the bill. Despite this, however, there are still some issues left.
Among these issues include tech giants like Meta and Google expressing concern that the bill will make the state of California hostile towards the development of AI. Similarly, Meta also called out the bill to be detrimental to research efforts. Lastly, OpenAI expressed that federal regulations would be more fitting rather than the SB 1047 which is more confusing.
The biggest issue related to this bill, however, is the danger it poses on open-source AI models. A lot of people in the IT support and services industry use these models to quickly advance AI and decrease risks. In the bill’s provisions, however, companies such as Meta can be held responsible to supervise the use of these models. Because of this, Senator Wiener had made adjustments, including changing the open-source models regulation standards.
Even with these oppositions, however, the bill still has a lot of prominent IT support and services industry supporters. This shows the varied opinions on the matter within the industry.
Here at EBSolutions, we provide you with the technology to get ahead of your competitors. And yes, this includes AI-backed solutions for your business. We have a team of experts specializing in cybersecurity, IT support and services to elevate your business. Book a consultation today so we can get started.