Child-safety mobilisation against generative AI companions

Parents, educators, paediatric mental-health advocates, school administrators, and adolescent-safety organisations organise against a specific product category — relational generative-AI products marketed or accessible to minors (companion chatbots, character-AI apps, AI romantic partners, AI-driven social feeds targeting adolescents). The tactics include product-liability litigation on behalf of harmed families, testimony to state attorneys general and consumer-protection agencies, coordinated Congressional-committee appearances by bereaved parents, school-district policies restricting or banning specific products, coalition partnership with age-appropriate-design-code campaigners, and media strategy that centres individual family stories to make the harm legible.

An actor chooses this strategy because generative-AI companion products entered mass adolescent use in 2023–2025 faster than any protective apparatus was built, and the constituency most exposed (adolescents) has no legislative voice, while the constituency with the strongest civic authority to speak on their behalf — parents, educators, paediatricians — has a long precedent of successful protective mobilisation on adjacent technologies (television violence, social-media harm, tobacco marketing to minors, smartphone-in-schools). The strategy converts a systemic-harm argument into a specific-product-line target on which regulators, prosecutors, and school administrators can act inside their existing authorities. It also imports a moral register — child protection — that is politically load-bearing across partisan divides in ways the broader AI-safety debate is not.

It trades issue-focus for movement-integration risk. The child-safety register is powerful precisely because it isolates a specific harm and constituency from the broader AI-critical field, but that same isolation makes coalition-work with the broader movement difficult — child-safety coalitions have historically produced regulatory wins the broader movement did not benefit from and vice versa. The strategy is vulnerable to being annexed by an adjacent political programme (state-level content-moderation legislation with speech-restrictive spillovers; parental-rights framings hostile to LGBTQ+ youth) whose priors are orthogonal to the AI-good movement. And the product-category focus can produce a whack-a-mole dynamic — a targeted vendor rebrands or a new vendor enters the category — without addressing the underlying design-incentive structure that produced the harm.

Ecology

Distinct from youth-led AI mobilization by voice: this strategy is for youth by adults with civic authority to speak on their behalf; youth-led mobilisation is by youth as the organising subject on their own generational fight. Distinct from AI literacy curriculum organizing in schools and libraries by register: literacy organising is proactive-educational, this strategy is protective-restrictive against a named product class. The two together — literacy plus protective restriction — is the fuller shape of the school-and-family-side response, but the coalitions that run each rarely overlap fully; the protective register carries constituencies (paediatricians, bereaved parents) that the literacy register does not, and vice versa.