AI-powered influence network
Anthropic said it detected and stopped a UAE-directed influence operation that aimed to weaken the Muslim Brotherhood worldwide. In its September intelligence threat report, the company described the effort as a coordinated campaign spanning the Atlantic and the region, run by what it called a single actor using its Claude AI assistant. The operation built a network of roughly 300 inauthentic influencer accounts on social media to try to shape discussion of the Muslim Brotherhood, the war in Sudan, and the United Nations.[S1]
Anthropic reported that the actor leveraged Claude to establish a fake NGO that appropriated the identity of an actual Swiss organization, then issued human rights reports authored by a state under that purloined name. The same AI assistant was used to draft official testimonies meant to be presented by two individuals at the UN Human Rights Council's 62nd session. Anthropic noted the material was crafted with deliberate constraints so that neither address would reference the UAE.[S1]
Targeting critics and accountability
Anthropic further stated that the campaign employed Claude to investigate and build profiles of 18 European Parliament members along with well-known journalists. It assembled counter-accountability files on UN Special Rapporteurs who had criticized the UAE's actions in Sudan. Although the operation was designed to evade detection, Anthropic said its probe tied it with high confidence to Emirati government officials, and that senior UAE officials were identified as the intended recipients of the material.[S1]
