California Senate Judiciary Committee, With bipartisan supportapproved Senate Bill 243 this month It calls for AI companies to “protect users from the addictive, isolated and influential aspects of artificial intelligence chatbots.” It’s the first bill of this kind in the US
On the day of the bill’s hearing, its author, Sen. Steve Padilla (D-Calif), held a press conference last year with Megan Garcia joining him. Sued AI Company Character.aiclaims that the chatbot played a role in his son’s suicide.
Garcia testified in support of the bill, and such chatbots say “It’s inherently dangerous“And it can lead to inappropriate conversations and self-harm. “Technological innovation is very important, but our kids cannot use it as guinea pigs to test the safety of new products.” Padilla said.
The same bill is currently underway Congresses of several states. These are important steps in the right direction. Equal laws are urgently needed nationwide.
What is at risk? Our children’s safety and emotional development, and the ability to critically think – perhaps even our democracy.
a 2024 Pew Research Poll Almost half of Americans reported using AI several times a week, and it turns out that one in four people use it “almost constantly.” a 2025 Gallup Survey It revealed that almost every American relies on products that contain AI.
New research is beginning to uncover important findings. Research in 2025 Published in society “We found a very strong negative correlation between subjects’ use of AI tools and their use of critical thinking skills.” Young users were particularly affected – trends Many teachers I’m starting to observe my students.
“As individuals increasingly offload cognitive tasks into AI tools.” Written by Michael GellichHe led a study that “decreases the ability to critically evaluate information, identify biases, and engage in reflexive reasoning.” If there was one, it’s a siren alert.
A far cry from perfect (or neutral) AI systems are built by humans, Intrinsic biaseven if it’s unintentional. Executives and developers from leading AI companies such as Openai, Google, and Meta will tweak chatbots and establish settings and rules. We rely on AI to think, so we outsource our personal ideas to the whims and prejudices of private companies and their teams.
Social media companies including Snap and Meta (owning Facebook, Instagram, WhatsApp and threads) are currently We develop our own “AI companions” from around the world. Billions of people, including hundreds of millions of children and teens, are offering “friends” that are always available online. While that may sound consolation, it deprives young people of the emotional growth and interpersonal skills they need for a real relationship.
Our AI companions are programmed to monetize our relationships by pretending to be trustworthy friends. While they are programmed Mines, records, and cleverly analyzed Everything we say and enter. Like a high-tech tutletail, they put this data into Data Ecosystemallowing marketers, advertisers and others to pay to target and manipulate us more than ever before.
January, Meta has been announced These chatbots are drawn from user interests, posts, and even dietary preferences and programmed using personalized “memories.” As Millions of Americans Pity emotional, political and developed Even sexual attachment Catapult with fellow AI members, despite their promises, suffering and misfortune. March, Research on MIT Media Lab Openai found that frequent use of AI chatbots correlates with “increased loneliness, emotional dependence, and reduced social interactions.”
As we are increasingly relying on AI to understand the world, we open ourselves up to manipulation by entities that do not have our best interests in mind. NewsGuard, 2025 news review service A serious threat has been revealed To AI Systems: A foreign disinformation campaign targeting AI training data with intentional falsehood. The Russian-related Pravda network published 3.6 million articles in 2024 designed to manipulate AI responses. Spread propaganda.
What happens to democracy when you offload your thinking to a chatbot that is being actively manipulated by foreign enemies seeking confusion?
ai stays here. And it could improve our lives in amazing ways Sclerotic disease In End poverty To achieve Scientific breakthrough And more. There are a few important steps to take right now, rather than other ways to make AI useful to us.
First of all, transparency is paramount. Through voluntary or legislative missions, large AI and social media companies such as Meta, Google, and Openai need to disclose the data they collect from us and who share it.
Food nutrition labels can help you make healthy choices by telling you whether you have high sugar or cholesterol. Similarly, aiNutrition label” We know whether AI systems are known to have a large amount of political bias, or how well they protect their privacy. Importantly, businesses can provide everyone with the ability to opt out of manipulative personalization.
Second, new regulations are needed to protect children, teens and users of all ages from the threat posed by “AI Companions.” Laws like California Senate Bill 243 It can prevent AI chatbots from adopting addictive engagement techniques and delegating protocols to handle pain and suicide signs. This type of targeting law deserves national adoption.
Third, new media literacy initiatives are essential. Research shows Teaching students how to find information can reduce the impact. Some state legislatures It has already moved in this direction, and has adopted media literacy as a standard teaching method for K-12 students. Skills including critical thinking and media literacy in the age of AI should be essential for students across the country, just like reading and mathematics.
The AI is a powerful sword with sharp sides. It can be used responsibly, keep children responsible, children independent, reasoning organized, and the ability to communicate authentically.
Mark Weinstein He is a ideological leader in technology, privacy expert and one of the inventors of social networking. he”Online sanity restoration: An innovative social framework” (Wiley, 2025).
