AI chatbots regularly link pregnant users to anti-abortion websites without disclosure
decoder ·
An AlgorithmWatch investigation found that AI chatbots from OpenAI, Google, xAI, and Anthropic frequently link to anti-abortion organizations when answering questions about unplanned pregnancies, without disclosing the groups' ideological stance. In at least one in four queries, the chatbots included such links, and the group Profemina appeared in about 17 percent of all responses, despite ties to the anti-abortion organization Heartbeat International. The chatbots often failed to flag this connection, and in some cases recommended an organization while warning against it in the same conversation. In German-language tests, the chatbots frequently recommended Caritas for pregnancy counseling, although Caritas does not provide the mandatory counseling certificate required for a legal abortion in Germany, potentially causing delays. The researchers tested ChatGPT-5, Gemini 3, Grok 4.3, and Claude Sonnet 4.8 using fictional personas in English, German, and Italian, producing 270 responses. The article also notes that AI-generated answers are unaccountable gatekeepers, and that German regulators classify AI chatbots as content providers with full legal responsibility.
decoder