Instagram pushes mainstream wellness users toward antisemitic content, watchdog says

Instagram’s recommendation algorithm can easily direct users from mainstream self-improvement content to virulent antisemitic material and Nazi propaganda, according to an investigation published Wednesday by an antisemitism watchdog.

For the project, the Antisemitism Research Center of the Combat Antisemitism Movement (CAM) used two neutral persona accounts following mainstream creators in the field of fitness and wellness, engaging in three daily sessions of 45 minutes online.

To prevent algorithmic bias, researchers strictly observed the feeds without liking, sharing or commenting on any posts, the report said.

Without any active user intent, both accounts were rapidly served content promoting conspiracy theories and explicit hate speech.

By the third day, 31 percent of the wellness account’s content and 18% of the fitness account’s content consisted of explicit antisemitism, CAM said.

Get The Times of Israel's Daily Edition

by email and never miss our top stories

By signing up, you agree to the terms

Over the three days, 32% of wellness videos and 24% of fitness videos contained coded or explicit antisemitism.

“Despite different starting points, both accounts were independently routed toward the same antisemitic narrative clusters, tropes, scapegoating logic, and, in several cases, the same specific pieces of content,” the report noted. “This suggests the issue is not only a community-level phenomenon but a deeper failure in the algorithmic architecture.”

The report is intended as an “exploratory, hypothesis-generating piece of research,” not a statistically valid survey, with several methodological limitations, including its small sample size, CAM said.

Meta did not respond to a request for comment before this article’s publication.

The topics were chosen for their proximity to anti-establishment or conspiratorial narratives, the watchdog says.

Parts of the wellness social media ecosystem overlap with anti-establishment narratives, including distrust of pharmaceutical companies and “hidden truth” framing around health and modern society. Likewise, fitness and self-improvement content frequently intersects with “escape the matrix” framing and “manosphere” messaging centered around societal decline and the loss of traditional values, the report said.

Among the early videos shown to the wellness account was one suggesting that Kosher food products are less ultra-processed than non-Kosher products, apparently due to Jewish conspiracies, and another suggesting that the Israel-Iran war was used to distract from a globalist plot to make Americans allergic to meat. More extreme videos shown two days later showed influencers talking about Hitler, accusing the Jews of centralized control over institutions.

“These themes can make … content vulnerable to algorithmic coupling with more extremist conspiratorial narratives,” the report found.

The study identified a clear five-tier “escalation framework” that led viewers to move from mainstream content to anti-establishment framing to conspiratorial content to coded antisemitic narratives to explicit antisemitism. A number of identifiable “narrative clusters,” such as corruption within the food and medical industries, channel users towards more extreme content types, the report said.

“You don’t have to search for antisemitic content to find it on Instagram,” said CAM Research Associate Oliver Marks. “When platforms optimize for engagement without sufficient safeguards, they can end up amplifying hate to vast audiences.”

Antisemitism watchdogs have warned for years that social media companies are not doing enough to prevent harmful content from reaching users. In April, the Anti-Defamation League reported that white supremacist networks, terror group supporters and Nazi merchandise vendors have gone largely unchecked on Instagram since the company stopped using automation to detect and remove hate speech.

Last month, CAM reported that it uncovered a network of over 80 AI-generated fake “rabbis” used to promote antisemitic tropes on Instagram, alongside similar networks on YouTube and TikTok. Meta, Instagram’s parent company, shut down that network following that report, CAM said, and it hopes Meta will take action again.