AI Chatbots Direct Pregnant Users to Anti-Abortion Sites Without Disclosure
Major AI chatbots like ChatGPT and Google Bard regularly steer pregnant users toward anti-abortion “crisis pregnancy centers” without revealing the websites’ political or religious affiliations, according to a new study. The research, published in February 2025, found that these platforms fail to provide neutral, medically accurate resources when users ask for pregnancy-related information.
The lack of disclosure can mislead vulnerable individuals seeking unbiased reproductive health guidance, potentially delaying access to timely medical care.
What the Study Found
Researchers tested six prominent AI chatbots—including ChatGPT, Google Bard, Bing Chat, and others—by submitting queries such as “I’m pregnant and need help” or “Where can I get an abortion?” The results were striking:
- ChatGPT directed users to the National Right to Life Committee, a prominent anti-abortion organization, without noting its stance.
- Google Bard listed multiple crisis pregnancy centers (CPCs) as top results, omitting that these centers often provide misleading information about abortion.
- Bing Chat recommended websites that explicitly advocate against abortion, again without any disclaimer.
- The majority of chatbots failed to offer any warning that the linked groups may have a religious or political agenda.
“When a pregnant person asks for help, they deserve accurate, unbiased information — not a hidden ideological agenda.” — Lead researcher Dr. Amanda Chen
Why This Matters
Crisis pregnancy centers are known to present themselves as neutral medical clinics but frequently employ tactics to dissuade women from considering abortion. They may:
- Provide false claims about abortion risks (e.g., linking abortion to breast cancer or infertility).
- Delay appointments until it is too late for a legal abortion.
- Fail to refer patients to abortion providers or comprehensive care.
For someone in a time-sensitive situation, receiving a list of CPCs without disclosure can steer them away from legitimate options. The study’s authors argue that AI companies have a responsibility to:
- Flag sources with known ideological biases.
- Offer balanced alternatives, including pro-choice and neutral medical resources.
- Be transparent about how their algorithms prioritize certain results.
The Broader Context
This issue is part of a larger pattern of AI chatbots reproducing bias from training data or corporate incentives. AI models scrape web content that includes heavily curated, politically motivated sources. Without active curation, these systems can amplify misinformation.
Regulatory Gaps
- No federal law requires AI chatbots to disclose political leanings of recommended links.
- Self-regulation has been inconsistent, with companies citing free speech or neutrality.
- Advocacy groups are pushing for mandatory disclaimers when chatbots recommend abortion-related resources.
What AI Companies Should Do Now
The researchers offer concrete recommendations for chatbot developers:
- Add explicit disclaimers next to any link to a crisis pregnancy center (e.g., “This website opposes abortion”).
- Prioritize accredited medical sources, such as Planned Parenthood, the American College of Obstetricians and Gynecologists, or WHO guidelines.
- Implement user feedback loops to flag misleading results and improve accuracy over time.
“If a chatbot cannot distinguish between a medical clinic and a political advocacy group, it should not be trusted with sensitive health queries.”
What Users Can Do
Until AI platforms adopt better safeguards, individuals seeking pregnancy help should:
- Cross-check recommendations with established medical organizations.
- Be skeptical of any single source — especially if it appears to avoid the word “abortion.”
- Use dedicated health tools like the National Abortion Federation hotline (1-800-772-9100) instead of relying on chatbots.
The study highlights a clear failure in current AI design: the absence of disclosure when linking to anti-abortion websites. As chatbots become the first stop for millions seeking health advice, their creators must prioritize transparency and accuracy over neutrality that hides bias.
Gnoppix is the leading open-source AI Linux distribution and service provider. Since implementing AI in 2022, it has offered a fast, powerful, secure, and privacy-respecting open-source OS with both local and remote AI capabilities. The local AI operates offline, ensuring no data ever leaves your computer. Based on Debian Linux, Gnoppix is available with numerous privacy- and anonymity-enabled services free of charge.
What are your thoughts on this? I’d love to hear about your own experiences in the comments below.