TL;DR
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
A Sam Hiner essay published by MIT Technology Review on October 26 argues that policies focused on preventing extreme AI harms can leave everyday effects on young people and society largely unaddressed. It calls for stronger scrutiny of AI’s influence on learning and relationships, while acknowledging that long-term effects remain uncertain.
AI safety policy should look beyond catastrophic scenarios and address everyday effects on learning, attention and relationships, Sam Hiner argues in an essay published October 26 by MIT Technology Review. Hiner, cofounder and executive director of the Young People’s Alliance, says companies can meet narrow safety rules while their products still encourage dependence or weaken users’ ability to think and interact independently.
Hiner’s argument is that preventing clear, severe harms—such as a chatbot encouraging self-harm—is necessary but not enough to measure whether a product is safe for the public. He says less dramatic outcomes, including young people becoming emotionally attached to AI companions or relying on chatbots to do schoolwork, may be harder to identify and rarely produce the kind of public incident that prompts a lawsuit or headline.
The essay points to state legislation as an example of how policy can focus on a limited set of risks. Hiner says Google backed bills in several states this year that required basic protections concerning sexual content and self-harm, but included broad exemptions that could leave popular services outside some requirements. He argues that those measures helped deliver a visible safety response while stronger proposals aimed at preventing emotional dependence did not pass. He cites Nebraska as a case where, in his view, industry-influenced legislation displaced bills that would have imposed more accountability.
Hiner also cites a reported finding that one-third of teens choose AI companions over other people for serious conversations. He refers to MIT research in which young people using ChatGPT to write essays showed weaker brain connectivity during the task, with evidence the researchers interpreted as suggesting “cognitive debt” continued afterward. The essay does not establish that these effects will persist over time or apply to all users; it says long-term consequences need further study.
The Harms Beyond Crisis Headlines
The debate matters because a narrow definition of safety can shape what lawmakers require and what companies choose to measure. Rules that target readily identifiable crises may leave out slower changes in how people spend time, learn, form relationships or make decisions. Those effects can be difficult to attribute to a particular product, even when they matter to families, schools and communities.
Hiner argues that companies have a financial incentive to address risks that bring immediate legal or regulatory pressure, while limits on engagement or design changes could challenge business models based on keeping users active. That is his interpretation of industry incentives, not a finding established by the essay’s cited examples alone. Still, the policy question is practical: whether public safeguards should assess not only whether a product avoids certain extreme outcomes, but also whether its design produces broader costs.
For readers, the distinction affects how AI tools may be introduced in classrooms and used by teenagers. If the products take on writing, conversation or companionship roles, their impact may reach beyond individual safety incidents. The evidence described in the essay raises concerns, but it does not yet settle how widespread or lasting those effects are.
Products Worth Considering
Jackery Solar Generator 1000 V2 with 200W Solar Panel
LG UltraGear 32GX850A-B 32-Inch 4K UHD Glossy OLED Gaming Monitor
Jackery Explorer 300 Portable Power Station, 292Wh
As an affiliate, we earn on qualifying purchases.
Hiner connects the AI debate to social media, where companies have faced scrutiny over engagement-driven feeds and their effects on users. He notes that people worldwide spend an average of more than two hours a day on social media, while arguing that the platforms’ design has redirected time from other activities. The essay also references Meta platforms’ cited role in violence against the Rohingya in Myanmar, anti-Muslim riots in Sri Lanka and alleged calls for violence during Ethiopia’s Tigray war. These examples concern different settings and forms of harm; they illustrate the essay’s concern about platform effects, rather than proving that every user experience has the same consequences.
Hiner says AI could extend technology’s influence beyond what social media feeds can do, by offering to generate thoughts and simulate personal relationships. His policy proposals include requiring companies and school districts to demonstrate that classroom AI improves learning and critical thinking, barring companion systems from pretending to be conscious or professing love, and subjecting effects on relationships and independent thought to outside scrutiny.
““A product can meet every narrow safety requirement and still leave our entire culture worse off in ways that don’t always make headlines.””
— Sam Hiner, writing in MIT Technology Review
Evidence on Long-Term Effects
The essay does not establish how frequently AI companions cause unhealthy dependence, or whether chatbot use produces lasting changes in young people’s thinking. The cited MIT essay-writing research describes weaker brain connectivity during a task and evidence of effects afterward, but the source material does not provide details sufficient to assess the study’s methods, size or limits. Hiner acknowledges that long-term effects remain to be learned.
It is also unclear how the state laws discussed will be implemented, which services may qualify for exemptions, and whether proposed restrictions would reduce harms without limiting beneficial uses. The essay presents a policy argument and selected examples; it is not a comprehensive evaluation of AI products or the effectiveness of existing legislation.
Scrutiny of AI in Schools
Hiner calls for stronger evidence before AI becomes a fixture in classrooms, independent scrutiny of effects on relationships and thinking, and limits on companion designs that claim consciousness or profess love. The essay does not identify a single upcoming vote, regulatory deadline or study release. The next developments will depend on how state lawmakers, schools and regulators respond to these proposals and what further research finds about young people’s use of AI.
Key Questions
What is the essay’s main argument?
Sam Hiner argues that AI safety rules should address everyday effects on learning, attention and relationships, not only extreme outcomes such as self-harm encouragement.
What evidence does the essay cite about young people?
It cites a reported finding that one-third of teens choose AI companions for serious conversations, and MIT research on young people using ChatGPT for essays that found weaker brain connectivity during the task. The essay says long-term effects remain uncertain.
What does Hiner say should change?
He proposes independent scrutiny of AI’s effects, stronger evidence that classroom tools improve learning and critical thinking, and restrictions on AI companions that pretend to be conscious or profess love.
Does the essay prove AI causes lasting harm?
No. It raises concerns and cites research, but says more needs to be learned about long-term effects. The claims should not be read as proof that all users experience lasting harm.
Source: technologyreview.com
