Source: Reuters

OpenAI, a California-based company and developer of ChatGPT, is supporting a bill that would force technology companies to label Artificial Intelligence (AI)- generated content. This content can range from harmless memes to deep fakes intended to spread misinformation about political candidates.

The bill, dubbed AB 3211, has so far been overshadowed by the attention paid to another California state AI bill, SB 1047, which requires AI developers to perform safety testing on some of their models.

AB 3211 also proposes distinguishing between products made by humans and artificial intelligence, which is a step toward establishing the rules of the game in the artificial intelligence scenario. Likewise, companies such as Microsoft have been in favor of this bill.

That’s because, on the subject of AI, California state lawmakers attempted to introduce 65 AI-related bills this legislative season, according to the state’s legislative database. These bills included measures to ensure that all algorithmic decisions are unbiased and to protect the intellectual property of deceased persons from exploitation by AI companies.

In this sense, AB 3211 has already been approved by the state Assembly with 62 votes in favor and 0 against it. It was also approved by the Senate appropriations committee earlier last month, which is preparing it for a vote by the entire state Senate. If passed, it would require synthetic intelligence companies to place an invisible watermark on all content created or “significantly modified” by their models.

Finally, the bill project would also require that IA companies provide “watermark decoders” so users can quickly identify if the content is AI-generated.