https://gizmodo.com/californias-ag-tells-ai-companies-practically-everything-theyre-doing-might-be-illegal-2000555896 [p] Skip to content Search * Home * Latest * Tech * Reviews * Science * io9 * AI * Video * Deals * CES 2025 Awards Artificial Intelligence California's AG Tells AI Companies Practically Everything They're Doing Might Be Illegal According to a recent legal memo, Silicon Valley's hottest business may be entirely based around criminal activity. Lucas Ropek By Lucas Ropek Published January 28, 2025 | Comments (52) | X Copied! OpenAI CEO Sam Altman at an Apple event (c) Justin Sullivan/Getty Images The legality of the AI industry's business practices has long been a hanging question. As a "disruptive" new technology, artificial intelligence has caused a wealth of problems at the very same time that it has offered society new benefits. Notably, AI has been used to mislead consumers, to create new forms of disinformation and propaganda, and to discriminate against certain groups of people. Now, the California Attorney General's office has issued a legal memo emphasizing the fact that all of that stuff is probably illegal. On January 13th, California AG Rob Bonta issued two legal advisories that illustrate all of the myriad areas where the AI industry could be getting itself into trouble. "The AGO encourages the responsible use of AI in ways that are safe, ethical, and consistent with human dignity," the advisory says. "For AI systems to achieve their positive potential without doing harm, they must be developed and used ethically and legally," it continues, before dovetailing into the many ways in which AI companies could, potentially, be breaking the law. Some of those ways include: * Using AI to "foster or advance deception." If you hadn't noticed, the internet is currently awash in a veritable tsunami of fake content. Concerns about a new generation of deepfakes and disinformation have exploded ever since AI content generators became popular--and with good reason. California's memo makes clear that companies that use AI to create "deepfakes, chatbots, and voice clones that appear to represent people, events, and utterances that never existed" could fall under the category of "deceptive" and, thus, be considered a breach of state law. * Falsely advertising "the accuracy, quality, or utility of AI systems." There has been quite a lot of, shall we say, hyperbole, when it comes to the AI industry and what it claims it can accomplish versus what it can actually accomplish. Bonta's office says that, to steer clear of California's false advertising law, companies should refrain from "claiming that an AI system has a capability that it does not; representing that a system is completely powered by AI when humans are responsible for performing some of its functions; representing that humans are responsible for performing some of a system's functions when AI is responsible instead; or claiming without basis that a system is accurate, performs tasks better than a human would, has specified characteristics, meets industry or other standards, or is free from bias." * Create or sell an AI system or product that has "an adverse or disproportionate impact on members of a protected class, or create, reinforce, or perpetuate discrimination or segregation of members of a protected class." AI systems have been shown to integrate human bias into their algorithms, which is particularly disturbing when you consider that AI is now being used to vet people for housing and employment opportunities. Bonta's office notes that automated systems that have disparate impacts on different groups of people could run afoul of the state's anti-discrimination laws. Bonta's advisory also includes a list of recently passed regulations related to the AI industry. The fact that the advisory says that all of these activities "may" break the law seems to signal that companies should effectively sell-regulate, lest they stray into criminal territory and tempt the state to take action against them. Bonta's memo clearly illustrates what a legal clusterfuck the AI industry represents, though it doesn't even get around to mentioning U.S. copyright law, which is another legal gray area where AI companies are perpetually running into trouble. Currently, OpenAI is being sued by the New York Times, which has accused the company of breaking U.S. copyright law by using its articles to train its algorithms. AI companies have repeatedly been sued over this issue but, because AI's foray into content generation represents largely unsettled legal territory, none of those lawsuits have yet been successful. Artificial intelligenceCaliforniaOpenAI Daily Newsletter Get the best tech, science, and culture news in your inbox daily. Select [ ] News from the future, delivered to your present. Select [ ] Please select your desired newsletters and submit your email to upgrade your inbox. [ ] Sign me up Leave this field empty if you're human: [ ] You May Also Like * Image: Mitchell Leff/Getty Images) Tech NewsArtificial Intelligence There Will Be More AI Ads and Fewer Movie Trailers in the Super Bowl What's $8 million for a 30-second spot when you're lighting money on fire anyway? By AJ Dellinger Published January 29, 2025 * Deepseek Ai Illustrations Tech NewsArtificial Intelligence OpenAI Claims DeepSeek Plagiarized Its Plagiarism Machine The White House's AI Czar said that DeepSeek 'sucked out the knowledge' of ChatGPT. By Matthew Gault Published January 29, 2025 * Photo of Capitol Hill. Tech NewsArtificial Intelligence OpenAI Is Launching Specialized ChatGPT for Government Use The product is not approved for government use yet, but OpenAI of course hopes President Trump will speed things up. By Thomas Maxwell Published January 28, 2025 * * A computer screen with DeepSeek's corporate logo, a blue whale, next to OpenAI's corporate logo, a stylized spiral. Tech News What DeepSeek's AI Did That Everyone Else's Didn't The Chinese AI company didn't surpass OpenAI by going bigger or inventing new techniques. By Todd Feathers Published January 28, 2025 * Japan Toray Tech NewsNews Analysis Remember When Nanotech Was the Next Big Thing? As the market reckons with AI's future, it's good to remember that we've been through this boom and bust hype cycle with technology before. By Matthew Gault Published January 28, 2025 * Illustration Deepseek Chatgpt Tech NewsArtificial Intelligence Chinese AI DeepSeek Deep-Sixes OpenAI on the App Store, Stocks Tank OpenAI is getting Temu'd, Sheined to the dustbin of history. By Matthew Gault Published January 27, 2025 Latest news Robert F. Kennedy Jr. at his confirmation hearing Bernie Sanders to RFK Jr.: Do You Recognize This Anti-Vaxxer Baby Onesie? Donald Trump at an even on September 23, 2024 in Smithton, Pennsylvania. USDA Inspector General Who Refused to Leave Post Escorted From Office by Security Star Trek Section 31 Michelle Yeoh Phillipa Georgiou Section 31 Is Now the Lowest-Rated Star Trek Project on Rotten Tomatoes Fossilized fish vomit. Yep, That's 66-Million-Year-Old Vomit [17ed43fa59ed6d8ddf0c] Sony Finally Gets Rid of Pesky PlayStation Network Login Requirement for Its Biggest PC Games Shield 3 Agents of SHIELD and Agent Carter Inch Slightly Closer to MCU Canonization Nordvpn Account NordVPN Says Its New Protocol Can Circumvent VPN Blockers Image: Mitchell Leff/Getty Images) There Will Be More AI Ads and Fewer Movie Trailers in the Super Bowl Best of CES 2025 Awards Best of CES 2025 Awards Best of CES 2025 Awards Latest news Bernie Sanders to RFK Jr.: Do You Recognize This Anti-Vaxxer Baby Onesie? 1/29/2025, 5:05 pm Robert F. Kennedy Jr. at his confirmation hearing USDA Inspector General Who Refused to Leave Post Escorted From Office by Security 1/29/2025, 4:20 pm Donald Trump at an even on September 23, 2024 in Smithton, Pennsylvania. Section 31 Is Now the Lowest-Rated Star Trek Project on Rotten Tomatoes 1/29/2025, 4:15 pm Star Trek Section 31 Michelle Yeoh Phillipa Georgiou Yep, That's 66-Million-Year-Old Vomit 1/29/2025, 3:55 pm Fossilized fish vomit. Latest Reviews The Nvidia RTX 5080 Is the GPU Gamers Should Really Want 1/29/2025, 9:00 am Nvidia Geforce 5080 03 The Samsung Galaxy S25's 'Now Brief' Is Its Most Underwhelming Feature 1/28/2025, 4:38 pm Galaxys25ultra Morning Brief Hero Narwal Freo Z Ultra Review: Better Eyes, Weaker Muscles 1/28/2025, 6:00 am Narwal Freo Z Ultra Robot Vacuum Just How Good Is Audio Eraser on the Samsung Galaxy S25 Ultra? We Put It to an Ear-Splitting Test 1/27/2025, 12:53 pm Galaxys25 Audio Eraser * * * X * * * * Sitemap * Edicion ES * Edition FR * Edicao PT * Accessibility * Privacy Policy * Terms of Use * Advertising * Reprints & Permissions * Your Privacy Choices * How We Review Products We may earn a commission when you buy through links on our sites. (c)2025 GIZMODO USA LLC. All rights reserved. * Cyber Monday * Best VPN + Best Free VPN + Best Cheap VPN + NordVPN Review + ExpressVPN Review + PIA Review + ProtonVPN Review + Surfshark Review + Super Bowl 2025 + US Open 2024 * Best Cloud Storage * Best Web Hosting * About Gizmodo Mode Follow us * * * X * * * * Home * Latest * Tech * Reviews * Science * io9 * AI * Video * Deals * CES 2025 Awards * How to Tip Gizmodo * About Gizmodo * + Cyber Monday + Best VPN o Best Free VPN o Best Cheap VPN o NordVPN Review o ExpressVPN Review o PIA Review o ProtonVPN Review o Surfshark Review o Super Bowl 2025 o US Open 2024 + Best Cloud Storage + Best Web Hosting + About Gizmodo Mode Follow us * * * X * * * [ ] Cookies settings