As long as the AI is, people are alarming what it can do to us: rogue sprintillence, mass unemployment, or environmental ruin from the spread of data centers. But this week, it was revealed that another danger – that creates unhealthy bonds with AI, is pulling AI protection out of the educational range and cross -hairs of regulators.
It has been bubbling for a while. Last year, two high -profile cases have been filed against Character Dot AI and Openi, alleging that their models like a partner have played an important role in the suicide of two teenagers. A Study Through the US -based Common Sense Media, which was published in July, it was found that 72 % of young people used AI for companionship. Stories in well -known shops about “AI psychosis” highlight that endless conversations with chat boats can lead people down to deception spiral.
The effects of these stories are difficult to increase. To the public, they are proof that AI is not just incomplete, but a technology that is more harmful than a helper. If you suspect that this grief will be taken seriously by regulators and companies, then there will be three things that can change your mind this week.
A California law enforcement.
On Thursday, the California state legislature passed the first bill of its kind. This will require AI companies to include reminders for users who are minors about which they know that the answers are born. Companies will also need to have a single protocol to commit suicide and self -harm, and consumer conversations with consumer chat boats will need annual reports on examples of suicide theory. It was led by Democratic State Senator Steve Padella, who was approved with heavy bilateral cooperation, and is now awaiting the signature of Governor Gavin Newsom.
There are reasons for doubting the effects of the bill. It does not specify that companies should identify which consumers are minors, and when one is talking about suicide, many AI companies include references to crisis providers. (In the case of Adam Rhine, a young man whose survivors are sued, was involved in his conversation with Chat GPT before his death, but the Chatboat allegedly moved forward. Advise Anyway about suicide.)
Nevertheless, the AI models are the most important efforts to curb such as a partner, which are also working in other states. If this bill becomes a law, it will take a shock to the position that the Openi has achieved, which is why “the United States is clear, not the best, the state or the patch of local rules with the rules,” as the company’s chief global affairs officer, Chris Lhani, ” Is written LinkedIn on them last week.
Is the purpose of the Federal Trade Commission
On the same day, the Federal Trade Commission made an announcement Inquiry In seven companies, they want information about how they produce roles like partner, make money from engagement, measure the effects of their chat boats and the effects of more. Companies are Google, Instagram, Meta, Openi, Snap, X, and Character Technologies, Characteristics.
The White House now has a lot of and potentially illegal, political influence on the agency. In March, President Trump sacked his lonely Democratic Commissioner, Rebecca Slaughter. In July, a federal judge ruled that the firing IllegalBut The last week The US Supreme Court temporarily allowed firing.
“Trump is a top priority for Venice FTC to protect children online, and thus promote innovation in the key sectors of our economy,” FTC Chairman Andrew Ferguson said in a press release.
Right now, that’s’s just – an inquiry – but this process (depending on how the FTC makes its search) shows internal tasks how companies use their AI colleagues to keep users back again and again.
Sam Altman on suicide matters
Plus On the same day (a busy day for AI News), Tucker Carlson published an hour -long interview with Open CEO, Sam Altman. It covers a lot of land – the fight with Alon Musk, the Open military consumers, the conspiracy theories about the death of a former employee – but the Altman has also included the most clear comments about suicide matters since talks with AI.
Altman talked about “tensions between consumer freedom and privacy and protection of weak users” in such matters. But then he made something that I had never heard before.
“I think it would be very reasonable for us to say that young people are talking seriously about suicide, where we cannot contact parents, we call the authorities.” Said. “This will be a change.”
So where does all this go ahead? For now, it is clear that at least in the case of children who are harming AI’s company – companies will not be aware of their familiarity. They can no longer end their responsibility by leaning on privacy, personal, or “consumer choice”. The pressure to take a tough line with state laws, regulators and angry people is increasing.
But how will it look like? Politically, the left and right are now focusing on the loss of AI to children, but their solutions are different. To the right, the proposed solution is associated with it Wave Internet -age verification laws that have now been passed in more than 20 states. These are to protect children from adult content by defending “family values”. On the left, the anti -trust and consumer protection options are restored to keep the Big Tech accountable to be held accountable.
A consensus on this issue is much easier than a deal on treatment. As it stands, it seems that we will completely end with the state and the patch of local rules, against which the Open AI (and many others) have lobbyed against it.
For now, companies have to decide where to draw the lines. They need to decide things like: When consumers are moving towards harming themselves, will the chat boats remove the conversation, or will it spoil some people? Should they be licensed and regular like therapeutic, or should be treated like entertainment products with a warning? The uncertainty is due to a basic contradiction: companies have made chat boats to take care of humans, but they have postponed to prepare the standards and accountability we demand for real careers. The clock is ending now.
This story was actually published on AI in our weekly newsletter, algorithm. To get such stories in your inbox first, Sign up here.