Computer Scientist Combats Contraceptive Misinformation in Post-Dobbs Era

In the wake of the Supreme Court’s 2022 Dobbs decision, which dramatically restricted reproductive healthcare access across America, millions turned to social media platforms for guidance on birth control options. Instead of reliable medical information, many encountered viral “horror stories” and misleading claims, promoted by algorithms designed for engagement rather than accuracy.

A new research initiative at Stony Brook University aims to combat this growing problem through advanced computational methods. Computer scientist Ritwik Banerjee has secured a two-year grant from the Society of Family Planning’s Contraceptive Misinformation and Disinformation initiative to analyze how birth control misinformation spreads online.

Banerjee, an assistant professor of research in the Department of Computer Science, brings a unique perspective to the project. Unlike other researchers in the grant cohort, who primarily come from public health or medical backgrounds, his expertise in computer science allows him to examine not just the content of misinformation but also the technical mechanisms that amplify it across platforms.

“Computational tools for language analysis let us see the ecosystem, not just a small sample,” Banerjee explained. His team will analyze over one million social media posts about contraceptive health, using transformer-based language models to identify recurring themes, frames, and rhetorical patterns associated with misinformation.

The project’s methodology extends beyond simple content analysis. Researchers will design AI agents to simulate how different user demographics might encounter contraceptive information online—from teenagers researching potential side effects to adults navigating new state-level restrictions. This approach provides insight into how recommendation algorithms serve different content to different users, potentially creating information bubbles.

Another crucial aspect of the study involves tracking how misinformation migrates across platforms. The team will document how content originating on one platform, such as TikTok, can spread to discussion forums like Reddit and eventually distort medical facts throughout the broader online ecosystem.

For Banerjee, this work connects to broader concerns about health equity and information access. Drawing on his experience with cross-cultural health narratives, he recognizes that health-related stigma and misinformation disproportionately impact already underserved communities. The shift from traditional word-of-mouth misinformation to algorithmically curated content has accelerated this problem.

The research addresses a significant gap in digital misinformation studies. While vaccine misinformation has been extensively documented and analyzed, contraceptive misinformation has received comparatively little attention from data-driven researchers, despite its profound impact on reproductive health decisions.

The project aims to produce tangible resources by the end of the two-year grant period. Banerjee’s team plans to develop open-source natural language processing plugins that health departments can implement to monitor emerging myths about contraception. They will also create practical API guidelines that social media platforms could adopt to reduce the amplification of harmful content, along with policy benchmarks for holding recommendation algorithms accountable in matters of reproductive health.

“I want a world where a health department with limited resources can use what we build; these tools should not be just for universities or big tech companies,” Banerjee said.

The ultimate goal extends beyond academic understanding. Banerjee aims to equip public health practitioners with effective tools to identify, respond to, and mitigate contraceptive misinformation before it spreads, helping to ensure that people seeking reproductive health information online can find reliable guidance rather than potentially harmful myths.

As social media platforms continue to serve as primary information sources for health decisions, this research represents an important step toward ensuring that algorithmic recommendation systems prioritize accuracy over engagement when presenting sensitive medical information to vulnerable audiences.