Courtesy: The quint
A massive and unprecedented generation of AI works that created deepfake images and videos targeting Muslims and their identity, which included Muslims being dehumanised, disenfranchised, displaced and erased from cultural memory, and, despite 119 documented breaches of the Model Code of Conduct, including Assam Congress chief Gaurav Gogoi being portrayed as a Pakistani agent and Muslim sympathiser, the Election Commission shut its eyes to it ahead of the Assam Assembly elections.
The findings emerged in a report released on Tuesday by the Foundation Diaspora in Action for Human Rights and Democracy, which described the electoral contest in Assam as witnessing what it termed “the first industrialised artificial intelligence disinformation operation” in an Indian state election.
The report argued that the orchestrated campaign relied upon technologically sophisticated yet politically incendiary content designed to manipulate communal anxieties while simultaneously reshaping electoral narratives.
The advocacy group contended that the disinformation apparatus functioned as part of a broader strategy in which Muslims were subjected to layered marginalisation, as their civic participation was allegedly undermined, their cultural representation distorted, and their socio-political presence progressively delegitimised.
The report, which was released two days before polling, stated that researchers had identified 31 confirmed deepfakes targeting Gaurav Gogoi, and that these fabricated visual narratives sought to construct an image of him as aligned with Pakistan while also depicting him as excessively sympathetic to Muslim communities.
It further recorded 119 documented violations of the Model Code of Conduct, yet asserted that no formal action had been initiated by the Election Commission in any of these instances, while social media platforms neither removed the content nor applied labels identifying the posts as artificially generated.
Describing the emergence of what it called a “disinformation infrastructure”, the report alleged that an extensive architecture comprising synthetic imagery, manipulated videos and AI-generated communal content had been assembled ahead of the polls, and that 432 posts on Facebook and Instagram were categorised as “very likely” or “likely” to have been AI-generated.
These posts, according to the findings, accumulated 45.4 million views and more than one lakh likes, indicating the scale of digital dissemination, while a single Instagram account identified as “politooons” reportedly generated 40.2 million views through 102 AI-generated posts, accounting for 88 per cent of total engagement.
The document also referred to a social media clip uploaded and subsequently removed by the Assam BJP, which depicted Chief Minister Himanta Biswa Sarma symbolically firing at images of two Muslim men, and which combined what appeared to be authentic footage with AI-generated visual targets.
The report noted that Sarma later stated in an interview that the video was “correct”, although he suggested the figures should have been identified as Bangladeshis, a remark that the advocacy group interpreted as reflecting a calculated linguistic recalibration rather than a substantive shift in intent.
The advocacy group further alleged that this sophisticated machinery was bolstered by "voter roll purges" and "demographic engineering," tactics that Chief Minister Mamata Banerjee claims are currently being mirrored in poll-bound West Bengal to marginalise Matuas, Rajbanshis, and Muslim communities.