Using artificial intelligence for a sentinel system to monitor online contraceptive-related disinformation
Contraception
Awarded 2025
Contraceptive misinformation and disinformation
Maria Gallo, PhD
University of North Carolina at Chapel Hill
$150,000

Deliberate efforts to spread false information about contraception can involve established myths or new false messages. The rise of novel disinformation messages can be unpredictable because they are not based on science or evidence. Crisis pregnancy centers (CPCs) are known to disseminate false claims about contraception. We propose to use artificial intelligence (AI) to develop a sentinel system for the dissemination of contraceptive disinformation in the US. Specifically, we will use natural language processing (NLP) techniques to 1) identify the key disinformation messages that CPCs currently disseminate through their websites; 2) detect whether any of the disinformation messages are novel; 3) evaluate whether the messages vary by geography; and 4) identify the rhetoric employed on the CPCs websites to promote the credibility of messages. We propose to focus on CPC websites for surveillance of contraceptive disinformation because our earlier research using population-representative surveys of adult, reproductive-age women in five states found that ever attendance at a CPC was not rare (9%-17%) and showed little variation by sociodemographic factors. People often attend CPCs after searching online for abortion services. In contrast, exposure to contraceptive messages via social media might reach a more limited population as people tend to follow like-minded individuals, creating echo chambers where their beliefs are reinforced. Developing a sentinel system that employs AI to monitor CPC webpages will enable us to proactively address contraception disinformation and to improve strategies for combating the rhetoric used to promote false claims.