Meta removes Iran-linked accounts using AI for U.S. disinformation

Meta has taken down a network of Facebook and Instagram accounts connected to an Iran-based influence operation that deployed AI-generated content to target American audiences with political messaging, the company disclosed in a threat report shared exclusively with Axios.

The operation removed four Facebook accounts and 31 Instagram accounts linked to Iranian actors. The network had accumulated roughly 79,400 followers on Instagram alone, representing what Meta classified as "moderate" reach for a state-backed campaign.

Why it matters

This disruption illustrates how foreign adversaries are integrating generative AI into influence operations at scale. While the tactics themselves aren't groundbreaking, the engagement levels achieved by these inauthentic accounts signal that AI-enhanced disinformation can build substantial audiences when operators invest in appearing authentic.

How the operation worked

The Iranian operators created personas posing as everyday Americans, including activists, students, and graphic designers claiming to live in major U.S. cities such as Washington, D.C., San Diego, and Atlanta. These fake accounts shared memes expressing anti-Republican viewpoints and messaging focused on the Israel-Palestine conflict and immigration issues.

Meta confirmed that operators used AI to generate some of the content posted across the network, including meme creation. The accounts actively tagged real journalists and politicians in posts and sent direct messages to high-profile political figures and news outlets attempting to establish content collaborations. None of these outreach efforts succeeded, according to a Meta spokesperson.

Detection challenges

The Iranian actors took deliberate steps to obscure their origins by routing all traffic exclusively through proxy and hosting services located in the United States and Canada. This infrastructure choice made detecting the campaign's true source more difficult for Meta's security systems.

Meta shared information about the operation with U.S. law enforcement following the disruption.

The broader AI disinformation landscape

Meta noted that AI-generated content has become routine in online influence operations, with the company finding the technology's use in virtually every influence network it now disrupts. The company has been investing in AI technologies designed to detect behaviors associated with inauthentic accounts and remove them more quickly.

While state-backed campaigns have historically achieved low reach and engagement, this particular operation generated what Meta described as "meaningful but limited" engagement—a notable outcome for foreign influence efforts.

These details were first reported by Axios.