AI Deepfakes Cast a Shadow Over Assam's 2026 Assembly Polls
The 2026 assembly elections in Assam witnessed an unprecedented and disturbing trend: the deployment of artificial intelligence-generated deepfakes as political weapons. These synthetic media creations did not merely attempt to tarnish reputations; they posed a direct threat to the integrity of the electoral process itself, with the full truth of their impact to be revealed only on May 4 during the vote count.
A Landscape Haunted by Synthetic Manipulation
Although the polls have concluded, a palpable undercurrent of fear persists regarding such AI-fuelled campaigns. The political landscape of Assam remains haunted by concerns over covert manipulation and synthetic media, with many fearing the state may have previewed a potential "AI apocalypse" for democracy. The technology's reach was extensive, targeting not only Chief Minister Himanta Biswa Sarma and his family but also opposition figures, including 27-year-old debutant Congress alliance candidate Kunki Choudhury in Guwahati and AIUDF women's candidate Meenakshi Rahman in the Pakabetbari constituency.
The damage appears significant and potentially irreversible. The reputations of the debutant and AIUDF candidates have been scarred, Sarma's family has been dragged into controversy, and voters across the state have been left to grapple with confusing, manipulated realities disseminated during the critical campaign period.
Political Repercussions and Legal Threats
In a recent telecast interview with ABP News, Chief Minister Himanta Biswa Sarma vehemently dismissed allegations against him and his wife. He accused Congress leader Pawan Khera of using deepfake documents, created via AI, to target his wife's properties just days before polling—a move he claimed was designed to sway voters. Sarma asserted that these tactics had no effect on the BJP's electoral fortunes, stating that Assam's voters reject both Khera and any slurs against their leaders.
Sarma announced a significant legal move, revealing that in his individual capacity, he will file a Public Interest Litigation (PIL) against the Election Commission after the elections, seeking to ensure the "purity of democracy." He argued forcefully that because these allegations were made prior to the election, they could have altered the outcome. "Can the election be recalled if the allegations were found to be wrong after the counting?" he questioned.
He further contended that a national political party executed this with the explicit motive of influencing the election's result. "Then their registration should be cancelled if the allegations are proved to be fraud. The press conference with the fake document by Congress is a collateral damage that cannot be repaired by any other means," Sarma stated. He emphasized his intention to take this issue to a larger platform through his PIL, warning, "(This time) It has not affected me. I think before 2029 such deep fakes and AI will be used against BJP."
Historical Context and the Election Commission's Response
This is not an isolated incident. Two years prior, in May 2024, the Delhi High Court addressed a similar PIL filed by a group of advocates during the Lok Sabha campaign. The petition expressed grave concerns over the rapid spread of manipulated videos and their potential to mislead the public during national polls. However, the court placed its trust in the Election Commission of India (ECI), stating it could not devise a new policy mid-election. It directed the lawyers' organization to submit a representation to the ECI, which was instructed to decide expeditiously by May 6 given the urgency.
Responding to these mounting concerns, the ECI issued a crucial advisory on May 6, flagging the misuse of hyper-realistic synthetically generated information, including deepfakes depicting political leaders, as a serious threat to free and fair campaigning. The Commission stressed that such content contaminates the electoral level playing field, misleads stakeholders, and erodes voter trust. It referenced earlier guidelines from May 6, 2024, and an advisory dated January 16, 2025, which addressed the ethical use of social media and the labelling of synthetic content.
The ECI's directives to political parties, candidates, and campaign representatives were clear and specific:
- All AI-generated or synthetically altered campaign material must carry a clear label such as "AI Generated," "Digitally Enhanced," or "Synthetic Content."
- Each content item must disclose the name of the entity responsible for its generation in metadata or captions.
- No content may misrepresent the identity, appearance, or voice of any person without consent in a manner likely to mislead voters.
- Parties must maintain internal records of all AI-generated campaign materials, including creator details and timestamps, for verification when sought by the ECI.
Legal Actions and Allegations from the Ground
On the legal front, Sarma's wife filed a First Information Report (FIR) on April 5 at midnight against Pawan Khera at the Guwahati Crime Branch. The complaint was lodged under multiple provisions of the Bharatiya Nyaya Sanhita, including charges of cheating, forgery, defamation, and false statements in connection with elections. Since then, Assam Police has been actively searching for the Congress leader, whose transit anticipatory bail issued by the Telangana High Court was stayed by the Supreme Court on a recent Friday.
Insiders within the BJP maintain that Khera's documents had no tangible impact on the party's performance. "Khera's allegations against the chief minister and his wife have turned out to be a shoot-and-scoot tactic by Congress," a BJP party insider revealed, downplaying the effect of the controversy.
Opposition Candidates Speak Out Against AI Attacks
Similarly, AJP candidate Kunki Choudhury, who contested against BJP's Vijay Gupta in the Guwahati Central constituency, found herself at the center of a deepfake storm. She lodged a formal police complaint against alleged deepfake videos circulating on social media about her family, following campaign remarks by Chief Minister Sarma targeting her mother's food habits.
"These deep fake videos falsely depict me in a manner intended to harm my character, reputation, dignity and electoral prospects," Choudhury stated in her complaint. She alleged before the media that some of the social media users who uploaded the fake videos are associated with the BJP. Her party leadership asserted that these deepfakes were circulated to incite public misunderstanding and undermine electoral dignity.
AIUDF central committee general secretary Champak Kalita confirmed that their candidate, Meenakshi Rahman, also fell victim to deepfake videos during the campaigning period. "We cannot ignore new technologies like AI tools in our daily lives or even in stock markets. But, these tools are good as long as they are not used to attack or smear anyone's character. Political parties may take help of AI tools to highlight their work but not against rivals," Kalita remarked, highlighting the ethical dilemma posed by advancing technology.
The 2026 Assam elections have thus set a alarming precedent, demonstrating the potent and disruptive power of AI in the democratic process. As the state awaits final results, the debate over synthetic media, electoral integrity, and the need for robust regulatory frameworks has only just begun, with implications that will likely resonate far beyond Assam's borders in future electoral contests across India and the globe.
