An election monitoring report has flagged what it calls an "unprecedented convergence" of AI-driven disinformation, electoral exclusion, and state-backed narratives in the run-up to the 2026 Assam Assembly elections, targeting 'Miya' Muslim community and the opposition leader.
The report by Diaspora in Action For Human Rights and Democracy, published on Tuesday, said 432 AI-generated posts across Facebook and Instagram reached 45.4 million views - produced through a six-tier content factory with a combined network reach of 407.4 million followers.
According to the report, thirty-one confirmed deepfakes targeted Congress CM candidate Gaurav Gogoi, fabricating his identity as a Pakistani agent and Muslim sympathiser, distributed through official BJP accounts and a verified Cabinet Minister's handle (Pijush Hazarika; 97,700 views).
Six AI-generated videos targeted Gogoi's wife, Elizabeth Colburn, a private individual holding no public office, depicting intimate and communal scenarios. The Instagram account politooons generated 40.2 million views across 102 AI posts, accounting for 88 per cent of all views of AI content.
In one cited incident, an AI-generated video posted by a ruling party handle depicted the CM shooting Muslims. The video was later removed, but the Chief Minister reportedly acknowledged altering language in such content to avoid legal scrutiny.
The report also said four exclusionary operations ran simultaneously against Assam’s Muslim communities - dehumanisation, electoral exclusion, physical removal and cultural erasure.
As an electoral exclusion, CM Sarma publicly declared that "4 to 5 lakh Miya votes will be deleted" through the Special Revision process, and 2.43 lakh names were subsequently removed from Assam's voter rolls. Simultaneously, the delimitation exercise reduced Muslim-majority constituencies from approximately 35 to 20.
The study noted that the Election Commission of India took no enforcement action against 119 documented Model Code of Conduct breaches - 84 classified as HIGH severity under Section 123 of the Representation of the People Act - including 15 breaches by the sitting Chief Minister himself.
“Social media platforms executed zero content takedowns; Meta's announced AI labelling policy produced zero labels on 172 AI-flagged
posts during the MCC window,” reads the report.
“What makes the Assam model uniquely dangerous is that it is not staying in Assam. The techniques documented here - voter roll purges, delimitation as demographic engineering, AI-generated communal content at an industrial scale - are being nationalised,” it added.