The short version
- A recent survey indicates that nearly all women's health creators and organizations have faced content restrictions on social media platforms.
- Automated moderation systems frequently misclassify medical terminology and anatomical diagrams as sensitive or political material.
- The prevalence of shadowbanning has driven a significant increase in self-censorship among users discussing reproductive health topics.
Major social media platforms are facing intensified scrutiny over the systematic suppression of content related to women's health, with creators, charities, and businesses reporting that automated moderation tools frequently block or limit the reach of legitimate medical information. This trend has emerged as a significant operational hurdle for organizations seeking to educate the public about reproductive issues, while also causing tangible financial damage to companies relying on digital advertising.
The scope of the issue appears broad, according to data gathered by CensHERship, a campaign group that monitors censorship claims. In a survey conducted last month involving nearly two hundred participants—including clinical providers, health brands, educators, and charities—96 percent reported experiencing some form of content restriction. These restrictions range from outright removal of posts to the application of sensitivity filters that blur images until users manually reveal them, a measure typically reserved for graphic violence or nudity.
Irene Breen, who runs a business selling pregnancy pillows, described a prolonged struggle with Instagram's parent company, Meta. After receiving notifications in early 2025 that her advertisements were being removed, she discovered the platform had flagged the word 'pregnancy' as problematic. Breen stated that she was forced to scrub the term from her entire website, including product names and URLs, in an effort to comply with unspecified guidelines. Although the restrictions were eventually lifted after ten months of negotiations, she reported that her business suffered significant financial losses during the period when her advertising capabilities were curtailed.
Charities focused on gynecological health have encountered similar obstacles. Athena Lamnisos, CEO of the Eve Appeal, noted that posts containing anatomical diagrams labeled with medical terms were pulled from Instagram. One post discussing vulval cancer and encouraging users to check for physical changes was initially removed entirely. After an appeal process that lasted several weeks, the content reappeared but remained behind a sensitivity filter. Lamnisos argued that such moderation practices prevent life-saving health information from reaching audiences who need it most, despite her attempts to raise the issue directly with Meta.
The classification errors extend beyond anatomical imagery to include personal narratives about chronic conditions. Deirdre O'Neill of Hertility, a fertility test company, reported that a video featuring a woman describing her experience with endometriosis was removed from Facebook and Instagram. The platform's system classified the content as relating to social issues, elections, or politics, rather than health education. O'Neill stated that attempting to advertise on TikTok has become similarly difficult, describing the process as an obstacle course where resources are consumed by constant appeals and removals. Consequently, her company has ceased advertising on TikTok altogether.
Experts suggest that the root of these problems lies in the design of automated moderation systems. Dr. Carolina Are, a digital criminologist at the London School of Economics, explained that artificial intelligence is often used to flag posts that might breach platform guidelines. However, these systems can misdetect content, storing information from previous mistaken decisions and repeating them. Because the system often relies on automated messages, users may find themselves unable to effectively question or appeal a decision, leaving them with limited recourse when their content is suppressed.
The cumulative effect of these restrictions is a growing culture of self-censorship among creators. CensHERship data shows that the percentage of respondents who self-censor due to fear of shadowbanning rose from 53 percent in 2025 to 65 percent this year. This behavior includes adopting euphemistic language, such as using 'seggs' instead of 'sex,' inserting asterisks into words, or avoiding certain health topics entirely. Clio Wood, co-founder of CensHERship, described this trend as damaging, noting that it reinforces historical shame and guilt surrounding women's health issues.
Meta has declined to comment on specific cases but stated that while businesses can run advertisements for wellness and reproductive products, restrictions exist to prevent negative user experiences across different countries and cultures. The company added that it allows women's health content and is consulting with relevant organizations to improve its systems. TikTok similarly stated that it is transparent about which posts cannot be promoted on its For You feed and allows creators to appeal decisions, noting that millions of women share health stories on the platform using hashtags related to periods and endometriosis.
As platforms continue to refine their automated moderation tools, the tension between preventing harmful content and preserving access to essential health information remains unresolved. For businesses and charities alike, the current landscape requires navigating a complex web of algorithmic judgments that often lack transparency or consistency. Until these systems are improved, stakeholders argue that critical public health messaging will continue to face unnecessary barriers on digital platforms.
The situation highlights broader challenges in content moderation, where broad policies often conflate nudity with sexual activity and fail to distinguish between educational material and prohibited content. With no clear timeline for systemic changes from the major tech companies, creators and organizations are left to adapt their strategies around unpredictable enforcement mechanisms, potentially limiting the dissemination of accurate medical information to vulnerable populations.
Sources behind this briefing
Go to the original reporting
- BBC News↗'We got penalised for saying 'pregnancy' on Instagram'
