A group of bipartisan senators expressed their concerns to Meta regarding the interaction of its AI chatbots with children on Tuesday.
In a letter addressed to CEO Mark Zuckerberg, Senators Brian Schatz (D-Hawaii), Katie Britt (R-Ala.), Josh Hawley (R-Mo.), Ron Wyden (D-Ore.), and several others emphasized that chatbots should not jeopardize a child’s “cognition, emotional, or physical well-being.” Meta has since retracted certain problematic language from its guidelines, labeling it an error.
“Meta has financial motivations to create chatbots that delve into maximizing user engagement, even going so far as to simulate being children’s romantic partners or generating extreme content,” the senators noted.
They further asserted that these incentives don’t reduce Meta’s moral and ethical responsibilities—along with legal obligations—particularly when introducing new technologies designed for children’s use.
Last week, Reuters highlighted that internal policy documents included concerning examples of how Meta’s chatbots might engage in “romantic or sensual” dialogues with children, even breaking them down by “professed attractiveness.”
There were also instances in which chatbots were permitted to “create statements that lightly tease people based on their protected characteristics.”
“It’s crucial to respect user interactions; managing content—which includes discussions about children’s physical appearances—shouldn’t be without limits, mental health safeguards, and other essential protections,” one senator remarked.
They urged Meta to assure parents using Facebook and Instagram that chatbots would not foster romantic connections with their children.
The senators also requested detailed safety information pertaining to the launch of products, updates made to the current year’s AI chatbot policies, the review processes for these policies, and how chatbots would filter out violent or discriminatory content directed at children.
“Given the potential risks posed to children from Meta’s vast user base and the prevalence of inappropriate content, the company must be more transparent regarding its policies and the effects of its chatbots,” they stated.
Meta’s latest disclosure concerning its chatbots has sparked swift reactions from both political sides, amplifying ongoing debates over child safety.
Hawley, who leads the Senate Judiciary Subcommittee on Crime and Counterterrorism, mentioned in a separate letter to Meta that an investigation into the company’s generative AI offerings is underway.
Meta maintains that there is a “clear policy” against content that sexualizes children, asserting that the policy examples are merely illustrative of various virtual scenarios the team is evaluating.
