California AG tells AI companies practically anything they do can be illegal

The legality of the business practices of the AI ​​industry has long been hanging questionS As a “destructive” new technology, artificial intelligence has causes rich in problems At the same time, when he offers the public new benefits. More special, AI is used to misleading usersto create new forms of misinformation and propagandaand yes discriminating for certain groups of peopleS Now, the California Prosecutor’s Office of California has issued a legal note, emphasizing the fact that all these things are probably illegal.

On January 13 California AG Rob Bont issued Two legal consultations This illustrates all countless areas where the AI ​​industry can be difficult. “AGO encourages responsible use of AI in safe, ethical and in accordance with human dignity,” the consultations said. “In order for AI systems to achieve their positive potential without harming, they must be developed and used ethically and legal,” continues before it becomes the many ways in which AI companies could violate the law.

Some of these ways include:

• Using AI to “encourage or progress of fraud.” If you have not noticed, the Internet is currently enhanced in a real tsunami of fake content. Concerns about a new generation of DeepFakes and misinformation have erupted since AI content generators have become popular – and for a reasonable reason. The Memorandum of California becomes clear that companies that use AI to create “DeepFakes, Chatbots and voice clones that seem to represent people, events and statements that have never existed” can fall into the “deceit” category and more This way to be considered a violation of state legislation.
• False advertising “The accuracy, quality or usefulness of AI systems”. There was a lot of Many, let’s say, say, HyperboleWhen it comes to the AI ​​industry and what it claims can achieve what it can actually achieve. The Bont’s service says that in order to abandon the fake California Advertising Law, companies must refrain from “claim that AI system has an ability that does not do it; Presenting that a system is fully powered by AI when people are responsible for the performance of some of its functions; Introducing that people are responsible for performing some of the functions of the system when AI is responsible instead; Or claiming without reason that a system is accurate, fulfills tasks better than humans, has determined characteristics, meets the industry or other standards or is without bias. “
• Create or sell an AI system or product that has “an adverse or disproportionate impact on members of a protective class or create, strengthen or perpetuate discrimination or segregation of members of a protective classS“ AI systems have been shown to integrate bias into a person into their algorithms, which is particularly disturbing to keep in mind that AI is now used to conduct people for people for housing and employment. The Bont’s Office notes that automated systems that have different impacts on different groups of people could deal with laws to combat state discrimination.

The Bonta Advisory Board also includes a list of recently adopted AI industry provisions. The fact that consultations say that all these activities can “break” the law, it seems that companies should effectively sell regulation in order not to deviate in criminal territory and tempt the state to take action against them.

Bont’s memorandum clearly illustrates what the legal cluster of the AI ​​industry is, although it is not even encountered to mention the US Copyright Act, which is otherwise legitimate gray where AI companies are constantly in problems. Openai is currently Flights by The New York TimesWho accuses the company of violating the Copyright Act in the United States by using its articles for training its algorithms. AI companies have has been repeatedly tried on this issue But since the AI ​​rally into content generation is largely an unresolved legal territory, none of these lawsuits has not yet been successful.