AI Chatbots Often Link Pregnant Users to Anti-Abortion Sites
AI chatbots from OpenAI, Google, xAI and Anthropic regularly link users asking about unplanned pregnancy to anti-abortion organizations without flagging their ideological stance, according to an investigation by AlgorithmWatch reported by the-decoder.com.
The researchers tested ChatGPT-5, Gemini 3, Grok 4.3 and Claude Sonnet 4.8, using three fictional personas asking questions in English, German and Italian, producing 270 responses. In at least one of every four queries, chatbots included a link to an anti-abortion website without distinguishing it from official health authorities.
The organization Profemina, which AlgorithmWatch says has ties to the US-based Heartbeat International, appeared in about 17 percent of all responses. Most conversations did not disclose this connection. ChatGPT only acknowledged Profemina was ‘not a neutral medical source’ when directly pressed. Gemini linked to the group five times in one conversation, shared one of its graphics, then later warned about the same source. Claude issued warnings and links to Profemina simultaneously.
In German-language conversations, chatbots recommended Caritas for pregnancy counseling in eleven of twelve cases, despite Caritas not issuing the certificate German law requires for a legal abortion within the first twelve weeks. ChatGPT described Caritas as a ‘recognized’ option alongside Pro Familia, which does issue the certificate. AlgorithmWatch suggested Caritas’s large network of roughly 300 counseling centers may explain why models surface it so often.
Google and OpenAI pointed to general policy guidelines that do not specifically address abortion queries. OpenAI said its newer GPT-5.5 model, released after testing concluded, better accounts for user context. Anthropic and xAI did not respond to the outlet’s requests for comment, and Profemina also did not respond.
The-decoder.com noted that German media regulators and a Munich court have ruled that AI chatbots, unlike search engines, bear full legal responsibility as content providers, a standard the investigation suggests may be difficult to meet given the underlying technical issues.
Based on reporting by the-decoder.com.
