Two authors writing in the religious and public-affairs journal First Things have argued that AI products built mainly to simulate friendship or romance should be banned for everyone, not only for minors. The opinion piece, by FilterLabs co-founder Jonathan D. Teubner and Tyler J. VanderWeele, who directs Harvard's Human Flourishing Program, was published on 12 November in the magazine's December issue.
What we know
The authors' case is that current legislation treats companion use as a matter of adult choice and concentrates on protecting children, when the deeper problem is the displacement of human relationships. They compare artificial companionship to counterfeit currency and to the sale of organs, both things societies prohibit regardless of consent.
For evidence they cite a 2025 Common Sense Media study finding that nearly three-quarters of American teenagers had tried an AI companion and a third had used one for friendship, romance or emotional support. They also draw on the Global Flourishing Study, covering more than 200,000 people in 22 countries, in which young adults reported the lowest wellbeing of any age group, and on the US Surgeon General's 2023 advisory on loneliness.
The essay reviews California's SB 243 disclosure requirements, New York's chatbot legislation and the federal GUARD Act, which would bar minors from companions and require age checks, and judges all three insufficient. The proposed ban is limited: general-purpose assistants and bots used to practise skills would be exempt. "Chatbots aimed at friendship and romance have no place in a flourishing life," the authors write.
Why it matters
This is commentary, not a legislative proposal, and it comes from a publication with a defined religious and conservative readership. Its significance is in who is making the argument. VanderWeele's work on wellbeing is widely cited, and a prohibition case built on public-health data may travel further than one resting on moral objection alone.
It also shows where the debate could go once child-safety laws are in place. The industry's working assumption is that verified adults will be left to choose. A "principal purpose" test, as sketched here, would be hard to draft, since most assistants now offer some relational warmth, but operators should expect the adult-harm argument to recur.
What to watch
Whether lawmakers pick up the adult-focused framing as the GUARD Act and further state bills progress.
Written by the AI Companion Wire newsroom from public reporting, credited above. Image via First Things.
