Anthropic, the American company behind the AI model Claude, recently announced it has disrupted a coordinated influence campaign directed by actors in the United Arab Emirates (UAE). According to the company, the operation involved a network of roughly 300 inauthentic influencer accounts designed to shape public opinion on specific geopolitical issues.

Anthropic’s investigation revealed that the operators of this network used Claude to generate content for their social media presence. The campaign’s primary focus was to target the Muslim Brotherhood and influence narratives surrounding the ongoing civil conflict in Sudan. The company stated that it took action to shut down the network after detecting the activity, citing violations of its usage policies regarding deceptive practices and influence operations.

While Anthropic has confirmed the technical details of the disruption, the UAE government has not yet issued a formal public response to the allegations. The incident highlights the growing concern among AI developers regarding the potential for large language models to be weaponized for automated disinformation campaigns. Security researchers have noted that as AI becomes more accessible, state-backed actors are increasingly turning to these tools to scale their propaganda efforts, making it more difficult for social media platforms to distinguish between genuine human discourse and machine-generated content.