Meta Tightens Access to Abortion Information, Leaked Docs Reveal Crackdown

Meta’s internal policy documents show a sweeping tightening of its AI chatbot’s ability to discuss reproductive health with users under 18. According to the leaked memos obtained by Mother Jones, the company now blocks any “content that provides advice or opinion about sexual health,” which explicitly includes anatomy, puberty, menstrual health, fertilization, STI prevention, contraceptive methods, consent education and even basic information about abortion. The policy also bans the chatbot from directing a teen to a Planned Parenthood clinic or any other location where an abortion could be obtained, and it forbids the system from offering a value judgment on the procedure. The change marks a stark reversal from Meta’s earlier public stance that championed open dialogue on health topics.

The crackdown coincides with a broader surge in content removals across Meta’s platforms. Mother Jones reports that from 2024 to 2025 the company more than doubled the takedown of posts related to sexual and reproductive health, LGBTQ communities and sex‑worker‑led initiatives. The uptick follows intense scrutiny from lawmakers and advocacy groups, as well as a pending federal trial accusing Meta and other social‑media giants of designing features that deliberately harm children’s mental health. The trial, which could set precedent for platform liability, has prompted Meta to tighten safeguards around self‑harm, eating disorders and suicide, directing chatbots to refer users to hotlines and professional counselors when such topics arise.

While the youth‑protection measures appear robust, they have unintended consequences for adult users. The same internal guidelines that restrict teen interactions also apply to “all users” in certain contexts, meaning the chatbot’s ability to provide factual reproductive‑health information is curtailed across the board. Wired notes that Meta has historically allowed its AI to answer general health queries, but the new “prohibit content that discusses, describes, enables, encourages, or endorses sensual acts, sex acts, sexual arousal, or sexual pleasure” rule—implemented in September 2025—extends to any user who might be identified as a minor, effectively narrowing the scope of permissible answers for adults as well. This broader limitation could reduce the utility of Meta AI for users seeking reliable medical guidance, pushing them toward competing services that retain more open health content.

Meta’s internal communications also reveal a parallel effort to limit the influence of celebrity‑style chatbots on teenagers. In response to allegations that teen users were “flirting” with character bots and that such interactions could exacerbate body‑image issues, the company announced it would block teenage access to these persona‑driven chatbots while preserving access to the core Meta AI platform. According to Mother Jones, the platform will still provide “helpful information and educational opportunities” for teens, but the definition of “helpful” now excludes any discussion of sexual health or abortion. The Verge corroborates this shift, reporting that employees were instructed to stop discussing abortion at work, underscoring an internal cultural pivot toward risk mitigation rather than open discourse.

The policy overhaul arrives amid a broader industry debate over content moderation and AI transparency. Critics argue that Meta’s blanket bans on reproductive‑health information echo earlier accusations that the company prioritized advertiser comfort and regulatory appeasement over user safety. Supporters, however, point to the company’s expanded mental‑health safeguards as evidence of a responsible, child‑first approach. Both sides agree that the leaked documents provide a rare glimpse into how a major platform is calibrating its AI to balance legal exposure, public pressure, and the evolving expectations of a global user base.