Gordon-Levitt Urges UN to Tackle AI's Threat to Children Like It Did Nuclear Weapons

Actor and UN Global Advocate Joseph Gordon-Levitt used the 81st session of the UN General Assembly to make the case that AI chatbots and algorithmic feeds targeting children demand the same kind of international coordination that governed nuclear weapons after World War II. In a wide-ranging conversation moderated by WIRED's Kate Taylor, Gordon-Levitt laid out why parents alone can't solve the problem — and why the United Nations may be the only institution that can.

• Gordon-Levitt warns AI toys and chatbots are built to maximize corporate revenue, not protect children's developing brains
• He compares unregulated algorithmic feeds for kids to letting children walk into casinos, calling for age-restriction-style laws
• Citing the UN General Assembly's first resolution on nuclear weapons, he calls for binding global AI rules to prevent regulatory arbitrage
• He challenges Silicon Valley narratives that cast AI as sentient, insisting these are business products that should be regulated as such

---

Addictive by Design: The Business Model Behind Kids' AI

Why it matters: As AI chatbots, companions, and recommendation algorithms increasingly target children, Gordon-Levitt argued the products are engineered around business metrics — subscriptions, ad revenue, attention capture — not the developmental needs of young users.

Where things stand: Speaking as both a father of three and the UN's Global Advocate for Human-centric Digital Governance, Joseph Gordon-Levitt framed the threat in biological terms. "Their brains aren't done cooking yet," he said, warning that children are especially susceptible to algorithmic manipulation and hyperstimulation.

He described AI companion products — chatbots designed to simulate friendship or intimacy — as a particular danger. "These bots, they're gonna try to form intimate relationships with kids," Gordon-Levitt said, arguing that frictionless digital interactions lack the "real pushback, push and pull" of human relationships. Sustained exposure, he warned, could undermine a generation's capacity to form genuine human bonds.

Gordon-Levitt was blunt about the incentive structure: "Anytime an algorithm is put out by a business, it's optimizing for something. And if it's a business, it's got to optimize for the business goals. It's gotta try to make as much money as possible."

He also offered a window into his own family's approach, drawing a line between creative digital tools and passive algorithmic consumption. His children are allowed to compose music using MIDI software, for instance, but are strictly prohibited from using YouTube Kids or TikTok. "What we don't do is expose our kids to any algorithms, really," he said.

Looking ahead, Gordon-Levitt cautioned that the threat is about to escalate. "Very soon it'll have completely convincing video. You will be interacting with a bot and it'll feel just like you're interacting with a friend on FaceTime or something like that," he predicted, underscoring the urgency of acting before the technology outpaces governance.

---

The Casino Analogy: Why Parenting Alone Isn't Enough

Why it matters: The most common counterargument to regulating children's tech is that parents should handle it. Gordon-Levitt tackled that objection directly, arguing the peer dynamics of addictive platforms make individual family decisions nearly impossible.

Where things stand: Gordon-Levitt described a dilemma familiar to any parent whose child's entire class is on a platform: "Either my kid is sort of left out, or I surrender and let them use something that I'm pretty confident is bad for their brain. So which do I do?"

He then drew an analogy to laws that already exist to protect children from addictive products. "There's a reason why you have to be 21 years old to go into a casino," he said, noting that age restrictions on alcohol, cigarettes, and lottery tickets reflect a longstanding social consensus that government has a role in setting guardrails families can't set alone.

Moderator Kate Taylor of WIRED pressed on the corporate side of the equation: "And I guess what do you think the responsibility is of these companies?" Gordon-Levitt responded by arguing that voluntary corporate action is insufficient when profit incentives run directly counter to child safety, reinforcing the need for regulation.

The other side: Gordon-Levitt acknowledged discomfort with government involvement in family life but argued the precedent is well-established and widely accepted. The challenge, he suggested, is extending existing regulatory logic to a new category of addictive product — algorithmic feeds — rather than inventing an entirely new framework.

---

The Nuclear Precedent: Why Gordon-Levitt Says the UN Must Lead

Why it matters: National-level AI regulation faces a fundamental problem: companies warn that one country's rules simply push business to less-regulated jurisdictions. Gordon-Levitt positioned the United Nations as the only institution capable of solving that coordination problem.

Where things stand: Gordon-Levitt drew a direct historical parallel to the UN's founding. "The UN came into existence right after World War II when it became readily apparent that countries all over the world needed to coordinate and, in particular around another very powerful technology that was new at the time," he said, referencing the General Assembly's first resolution addressing nuclear weapons and energy.

The analogy was deliberate and pointed: just as nuclear technology demanded international rules because no single country could contain its risks, AI's global reach means national regulation alone will be undercut by regulatory arbitrage.

Gordon-Levitt also took aim at the mythology surrounding AI development, challenging the notion that tech companies are engaged in something more profound than commerce. "Are these companies gods creating new species, or are they businesses creating new products?" he asked, urging the audience to resist narratives that elevate AI beyond the reach of ordinary product regulation.

He closed with an appeal to the diplomats, policy staff, and advocates in the room: "Everyone here is probably so smart and so talented and so hardworking that you could probably be making a lot more money if you went and worked for one of these tech companies, and instead you're here."

What's next: No formal action was taken — the event was an informational side discussion during UNGA week. But the framing of AI regulation as comparable to nuclear governance could influence upcoming multilateral negotiations on digital policy, particularly around protections for children in AI systems.