OpenAI, Microsoft, and Adobe are officially supporting the AB-3211 expenses that may need tech firms to attribute AI-generated material in California.
Watermarks must be added to the data of photos, video, and audio tapes that were produced or tremendously modified by conceptual AI systems, according to the policy.
The costs will also need large social media websites, such as X and Instagram, to implement “easy-to-understand” brands to uploaded content that people can visit to see. Audio videos and music videos may be subject to physical labels.
California State Assembly member Buffy Wicks received a notice from OpenAI presiding over which Reuters had not seen it in its entirety. Similar words viewed by TechCrunch also featured Adobe and Microsoft backing the bill.
According to Reuters, OpenAI Chief Strategy Officer Jason Kwon wrote in his letter that “new systems and standards may help people understand the nature of the information they find online, and prevent misunderstandings between realistic AI-generated content and human-generated content.”
In an election year, transparency and requirements around origin are particularly crucial because encouraging artificial content can be produced to sway votes, according to Kwon’s letter. For instance, in early August, AI-generated photographs of Taylor Swift fans expressing their support for Donald Trump were shared on social media.
AB-3211 has been approved by the state Legislature and senatorial budget commission. Before Governor Gavin Newsom is advanced to implement or veto it before the end of September, a total state Senate vote will follow.
TechRepublic reached out to OpenAI, Microsoft, and Adobe for remark.
View: Apple Joins Voluntary U. S. Government Commitment to AI Safety
The act is criticized by an advocacy group.
Microsoft and Adobe are members of the Coalition for Content Provenance and Authority, an organization that creates requirements for the identification of AI-generated information. The party contributed to the development of C2PA metadata, a platform for DALL-E 3’s use of embedding credible metadata into online content.
But, the BSA, an advocacy group for the technology industry that includes Microsoft and Adobe, objected to AB-3211 in a letter to California’s Assembly Privacy and Consumer Protection Committee in April. It took problem with:
- The scope of information that must be watermarked.
- The requirement to provide resources to check watermarks.
- The “unworkable” alert and reporting method it proposes.
Since then, politicians have amended the bill many times to make it seem as though they have appeased the concerns of the businesses.
California managing numerous charges related to AI
AB-3211 is distinct from SB-1047, California’s questionable AI Act.
By imposing strict safety standards on designers, SB-1047 aims to stop AI designs from causing significant harm to humanity or financial loss. SB-1047 was approved by the California Appropriations Committee earlier this month, but it still needs to be approved by the state Assembly and Senate before it becomes laws.
Many of Silicon Valley, including OpenAI, Meta, Google, and Hugging Face, formally disparaged SB-1047. Elon Musk, however, took to X before this week to claim that he thinks the bill needs to be passed. He posted:” For over 20 years, I have been an activist for AI rules, only as we manage any product/technology that is a potential threat to the public”.
Additionally, two former OpenAI people wrote a notice to Newsom criticizing their previous employers for opposing SB-1047.
” OpenA I opposes even the extremely light-touch requirements in SB1047, most of which OpenAI claims to voluntarily commit to, raising questions about the strength of those commitments”, they wrote. OpenAI disagrees with the researchers ‘ “mischaracterization of]their ] position”, per Business Insider
AB-3211 and SB-1047 are only two of at least 65 AI-related costs that legislators aimed to offer in California this month, aiming to stabilize technological progress with buyer protection.
