Instagram's Profit-Driven Algorithm Amplifies Extremism, Igniting Moderation Crisis

Millions of Gen Z users are encountering extremist content—including Holocaust denial and antisemitic propaganda—on Instagram Reels, often adjacent to advertisements from major brands like JPMorgan Chase, the U.S. Army, and Porsche, according to a Fortune investigation. The platform's algorithm, designed to maximize engagement, appears to amplify conspiratorial and racist content, which is then monetized through ad placements and creator incentives.

The issue highlights a growing tension between platform profitability and content moderation. Instagram's recommendation system prioritizes content that drives watch time and shares, including extremist material disguised as humor or aesthetic content. For example, a reel from the antisemitic account @goyimclub recently appeared directly below a JPMorgan Chase ad, using a Holocaust-denial trope to mock the scale of the genocide. Another reel, from @gelnox.exe, implied a false historical narrative, linking Spain's prosperity to the expulsion of Jewish people.

Meta, which owns Instagram, has acknowledged the problem, stating it "doesn't want this kind of content on our platforms" and has taken steps to remove copies of violating material. However, Fortune's reporting found that such content remains visible, with extremist reels resurfacing minutes after Meta's official response. Advertisers—including JPMorgan, Nationwide Insurance, and the U.S. Army—have not publicly addressed the adjacency of their ads to extremist content, despite being contacted for comment.

The phenomenon is exacerbated by Instagram's monetization tools, which reward creators for high-engagement content, regardless of its harmful nature. This has created a lucrative ecosystem for "clip-farm networks" that produce extremist material, further entrenching radicalization pathways among younger users. Experts warn that the algorithm's lack of nuance in distinguishing between educational content and propaganda leaves Gen Z vulnerable to manipulation.