Instagram's search remains a haven of vaccine misinformation, a year after Facebook said it wouldn't show such content on Instagram Explore or hashtag pages
Jesselyn Cook / HuffPost : Tweets: @jessreports , @mosseri , and @tonyajoriley Tweets: Jesselyn Cook / @jessreports : Respectfully, people *do* search for vaccine info on Instagram. There's a very active anti-vax community on the platform that spreads misinformation & benefits from IG's algorithmic amplification: https://www.huffpost.com/... https://twitter.com/... Adam Mosseri / @mosseri : @donie The main thing is that people don't search that much on Instagram, so we've prioritized our wellbeing work, vaccine related and otherwise, elsewhere. We should do more work here, but need to hold ourselves accountable for having real impact. Tonya Riley / @tonyajoriley : “It's been almost a year since Instagram pledged to reduce the spread of vaccine-related misinformation on its platform. But today, it continues to do the exact opposite,” @JessReports https://www.huffpost.com/...
Context & Ripple Effects
The March 2019 pledge to demote anti-vaccine Pages, halt recommendations, and reject ads was Facebook's flagship health-misinformation response, and Instagram was explicitly included via Explore and hashtag pages. Yet within two months CNN found anti-vaccination posts and a #vaccineskill tag still ranking in Instagram search, and The Atlantic flagged that Instagram had largely escaped scrutiny while its recommendation algorithm boosted conspiracy content.
A year on, HuffPost's reporting and Adam Mosseri's own reply — that Instagram users 'don't search that much,' so wellbeing work was prioritized elsewhere — concede the gap while reframing it as a traffic question. The exchange matters because it shows enforcement tracking where visibility is highest, not where misinformation actually lives.
First-order effects
- Facebook's health-misinformation pledge is now verifiably incomplete: the anti-vax community on Instagram continues to gain reach through search and algorithmic amplification a full year after the policy was announced.
- Mosseri's public admission that Instagram search was deprioritized hands reporters and regulators a concrete, quotable enforcement gap tied to a named executive.
Second-order effects
- Scrutiny migrates to the surfaces the pledge skipped: subsequent reporting found misinformation still dominating Instagram's search results months later, forcing Facebook into feature-by-feature fixes rather than a single policy fix.
- The pressure eventually produces blunt product interventions — by mid-2021 Facebook was hiding the #VaccinesKill tag from search entirely behind a 'keeping our community safe' message, two years after Instagram hid the tag — a sign that ranking tweaks alone were not holding.
Third-order effects
- If the pattern holds, platform health pledges get enforced where press and regulator attention concentrates (feeds, Explore, ads), while search and recommendation surfaces lag until each is separately exposed — moderation becoming reactive patchwork across features rather than a graph-wide policy.
- That patchwork dynamic invites regulators to treat search and recommender outputs as auditable products in their own right, since the company's own executives concede these surfaces are known to carry the content.
The trend: Social platforms are implementing misinformation pledges surface-by-surface, with search and recommendation features lagging flagship feed policies until sustained press scrutiny forces each fix.