Anthropic bans 29 AI accounts linked to Awami League

AI company Anthropic has shut down 29 AI accounts linked to an automated influence network in Bangladesh that allegedly exploited its chatbot to manufacture fabricated Bengali-language news targeting rural audiences.

The network, reportedly orchestrated by a single individual operating from Gaibandha district, ran for approximately 16 months, rotating between accounts to circumvent platform restrictions and evade detection, while consistently promoting narratives favourable to the Awami League and disparaging its political rivals. According to Anthropic's September 2026 report, the operation generated no fewer than 1,500 headlines, 300 fabricated reports and 1,500 image-generation prompts intended for dissemination across Facebook, YouTube and TikTok. The content allegedly propagated false claims targeting the BNP, Jamaat-e-Islami, the National Citizen Party and student leaders, including baseless accusations of conspiracies, foreign entanglements and corruption.

Read More: Apple unveils foldable iPhone Duo at $1,999

Anthropic stated it uncovered no evidence of state direction or funding behind the operation, notwithstanding the fact that certain narratives coincided with pro-India geopolitical interests. Given that the Awami League has remained out of power since the July 2024 uprising, the company noted that the network effectively operated on behalf of an opposition party rather than a sitting government. The company has since terminated all implicated accounts, fortified its detection mechanisms, and shared relevant evidence with affected social media platforms and other partners, although it acknowledged that disrupting the accounts does not necessarily guarantee the cessation of the underlying activity.