In a recent statement, OpenAI supports the California AI bill requiring the ‘watermarking’ of synthetic content. OpenAI, a leading developer in the artificial intelligence (AI) field, is supporting a California bill, AB 3211, which mandates tech companies to label AI-generated content. This initiative is aimed at increasing transparency, especially during election seasons, as AI-generated content ranges from harmless memes to more serious deepfakes that could spread misinformation about political candidates.
The proposed legislation, authored by California State Assemblymember Buffy Wicks, requires that all AI-generated content be labeled to help users distinguish between human-created and AI-created materials. The bill has passed the state Assembly with a unanimous 62-0 vote and cleared the Senate Appropriations Committee. A vote by the full state Senate is expected by the end of the legislative session on August 31. If passed, it would move to Governor Gavin Newsom, who has until September 30 to sign or veto it.
Concerns Over Misinformation
OpenAI supports the California AI bill requiring the ‘watermarking’ of synthetic content, particularly in light of its potential impact on elections and public opinion. As countries worldwide prepare for elections, the concern over AI’s potential to influence public opinion grows. OpenAI emphasized the importance of transparency in its communication with Assemblymember Wicks. “New technologies and standards can help people understand the origin of content they find online,” said OpenAI Chief Strategy Officer Jason Kwon. This push for clarity is particularly important given the increasing use of AI-generated content in elections, as seen recently in Indonesia.
While AB 3211 has garnered support, another California AI bill, SB 1047, has faced significant backlash. Proposed by State Senator Scott Wiener, SB 1047 requires safety testing for advanced AI models costing over $100 million to develop or use significant computing power. This bill also mandates developers to outline methods to deactivate AI models if they malfunction and threaten critical systems, such as the power grid.
Key Provisions of SB 1047
The bill empowers the state attorney general to take legal action against non-compliant AI developers, especially if the technology poses an ongoing threat. Additionally, the legislation calls for third-party audits of safety practices and provides protections for whistleblowers exposing AI abuses.
Despite revisions to SB 1047, including removing criminal penalties and eliminating a government oversight committee, the bill has met with resistance from the tech industry. Companies like Google, Meta, and OpenAI have raised concerns that the bill could hinder AI innovation in California. Meta, in particular, warned that the bill could discourage AI development in the state. OpenAI, in a letter to Senator Wiener, argued that the legislation could create an uncertain legal environment and drive AI developers and engineers away from California.
While SB 1047 has passed the state Senate and the Assembly Appropriations Committee, it still faces significant opposition. Notable opponents include California Congressional Democrats, such as Nancy Pelosi and Ro Khanna, who argue that the bill is “ill-informed” and could stifle technological innovation. They believe it could drive developers out of California and threaten open-source AI models, which rely on freely available code that anyone can use or modify.
Support from AI Pioneers
Despite the opposition, some prominent figures in the AI community have voiced their support for SB 1047. AI pioneers like Geoffrey Hinton, Daniel Kokotajlo, and Yoshua Bengio have endorsed the bill, emphasizing the need for safety and regulatory measures in AI development to prevent potential risks.
As OpenAI supports the California AI bill requiring the ‘watermarking’ of synthetic content, it aims to foster a more transparent digital ecosystem for all users. While some argue for stronger safety measures and transparency, others fear the impact of stringent regulations on innovation and growth in the AI sector. The outcome of these legislative efforts could set a precedent for AI governance in California and beyond.
Also Read: Amazon CEO Andy Jassy Says Gen AI Saved $260 Million and 4,500 Developer Years.
