Anthropic says UAE‑linked AI campaign targeted UN officials over Sudan

Anthropic announced on Thursday that it had identified and disrupted a sophisticated disinformation campaign it linked with high confidence to officials in the United Arab Emirates. The operation employed the company’s Claude chatbot to generate reports, advocacy material, social‑media posts and even forged testimonies for the United Nations, focusing on the Muslim Brotherhood and UN special rapporteurs critical of the UAE’s role in Sudan.

How the AI‑driven scheme worked

The campaign, centred on an AI persona called “Deadshot”, followed a doctrine described as a “coordinated transatlantic and regional operation to dismantle the Muslim Brotherhood globally”. It produced what Anthropic called “counter‑accountability dossiers” on UN special rapporteurs and fabricated the identity of a genuine Sudanese human‑rights group to draft testimony for two individuals. The goal was to make material appear to come from independent local witnesses rather than state‑sponsored messaging.

Anthropic estimates the operation ran roughly 300 fake social‑media accounts. It could not confirm whether the forged dossiers or testimonies reached their intended UN audiences or influenced policy decisions.

Why the UN and the wider world should care

Since April 2023, Sudan has been engulfed in a war between the national army and the paramilitary Rapid Support Forces (RSF), a conflict the United Nations has called one of the world’s worst humanitarian crises. The UAE has repeatedly been accused, including by UN experts, of supporting the RSF, a claim Abu Dhabi denies. If AI tools are used to mask state involvement, it becomes harder for the UN and NGOs to verify the provenance of information, potentially skewing diplomatic responses.

"Inaction means more suffering and death," US deputy ambassador Jeffrey Bartos said during a recent Security Council meeting.

The UN’s independent fact‑finding mission on Sudan recently highlighted that private actors and recruiters operating from the UAE and other countries were using transit points in Chad, Libya and Somalia to funnel personnel, training, weapons and logistics to the RSF.

What may happen next

Anthropic’s disclosure comes as the UN Security Council debates extending the arms embargo that currently covers western Darfur. The United States argues the embargo is “outdated” and insufficient, while Russia, a permanent Council member aligned with Sudan’s army, opposes any broadening of sanctions.

Should the UN adopt stricter verification protocols for AI‑generated content, it could curb the influence of such covert campaigns. Meanwhile, the UAE has not responded to requests for comment, leaving the international community to monitor the situation closely.

For more on Anthropic’s stance on AI safety, see Anthropic defends safety record as AI risk warnings surface.