BOOM Report: AI-led Disinfo, Media Misreporting Hit A Peak Between Aug-Oct 2026

AI-generated or AI-edited content featured in 65 of the 165 fact-checks BOOM published between August and October 2026, nearly four in ten. Mainstream news outlets had reported the claims in 13 of them. September was the first month this year in which AI made up the majority of fact-checks (26 of 48).

Between 3 August and 30 October 2026, BOOM published 165 original fact-checks in English (80), Hindi (67) and Bangla (18).

BOOM's fact-checks over these three months were shaped by false claims around the Nepal floods in late August, the BRICS summit in Delhi in September and a second wave of Cockroach Janta Party (CJP) protests against Chief Election Commissioner (CEC) Gyanesh Kumar, which built through October. In the last week of October, claims about India-China face-offs in Arunachal Pradesh led to several more.

AI-Related Fact-Checks: A September Peak, And Not Only Political

65 of BOOM's 165 fact-checks (39.4%) involved AI-generated or AI-manipulated content, up from 30.7% in January–July and 20.5% in 2025. By month, the AI share was 30.4% in August (21 of 69), 54.2% in September (26 of 48) and 37.5% in October (18 of 48).

What kind of AI

Synthetic AI images were the most common type, with 21 of the 65 AI fact-checks. Synthetic AI video and voice clones over real footage accounted for 12 each, AI-edited real photos and videos for 10, and face-swap or lip-sync deepfakes for 6. The remaining 4 were AI-made graphics, such as a fabricated BBC headline about India's Gen Z and a graphic claiming actor Rajinikanth had been appointed Kerala governor.

Deepfakes and voice clones together made up 18 of the 65 (27.7%), against 42.6% in January–July.

Where AI shows up

44 of the 65 AI fact-checks (67.7%) were political, down from 82.1% in January–July. The rest included a fake theft story claiming an "AI retro trend" of the 1980s helped catch a woman in Ujjain and an AI video passed off as an eruption of Indonesia's Anak Krakatau volcano.

AI was most common in Entertainment fact-checks, appearing in 9 of the 14, against 30 of the 84 Indian Politics fact-checks (35.7%). No communal fact-check involved AI.

Chatbot answers as evidence

AI chatbot answers were used to back false claims, which did not appear in BOOM's January–July fact-checks. Grok and Gemini wrongly told users that real photos from the CEC protests were not genuine, and NDTV ran a fact-check video on that basis before deleting it.

In another case, Grok and Google's AI Overview wrongly identified real photos of the crowd at an Air Force show in Varanasi as pictures from the 2025 Kumbh Mela. The false claim that BJP had used an old Kumbh photo to show a crowd for Modi spread widely, including through a repost with a mocking caption by AAP's Assam handle.

AI reached newsrooms

In 13 of the 32 fact-checks in which a mainstream outlet carried the false claim (40.6%), the content was AI-generated or AI-edited, up from 7 of 37 (18.9%) in January–July. News18 carried AI-manipulated images of actress Nayanthara with a shaved head, Hindustan Times carried the fake Ujjain theft story about an "AI retro trend", and NDTV Profit and The Times of India carried the AI video of the Anak Krakatau eruption.

Themes, Topics And Targets: What Drove The Fact-Checks

Indian Politics made up 84 of 165 fact-checks (50.9%), up from 42.3% in January–July, while International Politics fell to 23 (13.9%) from 26.2%. Entertainment rose to 14 (8.5%) from 4.4%.

Rahul Gandhi and Narendra Modi were the most frequent primary targets, followed by Delhi Police. The claims about Gandhi included a fake statement attributed to the Italian Embassy criticising him, carried by Republic, Dainik Bhaskar English and Gujarat Samachar. A photo of Mallikarjun Kharge carrying a "suitcase" was shared by BJP's Suresh Nakhua.

The second Cockroach Janta Party wave

The anti-CEC protests were the largest single story, with 24 fact-checks, 22 of them in October. In the 12 days from 5 to 16 October, 19 of the 22 fact-checks BOOM published were about them.

Half of the 24 used old footage. Visuals from the 20 July march were reshared as the 10 October protest, and old footage of a farmers' protest was shared as deaths at Jantar Mantar. A video from West Bengal was shared by Shiv Sena (UBT) MP Sanjay Raut as Delhi Police transporting uniforms, and was carried by Bangladesh's Ittefaq.

Fabrications circulated about both the movement and its critics. AI was used to make a surety bond bearing the signature of CJP founder Abhijeet Dipke, shared by TMC spokesperson Riju Dutta, and a video of NSA Ajit Doval threatening to leak nude photos of CJP leaders. CJP spokespersons Saurav Das and Ashutosh Ranka shared a fake quote attributed to Gaurav Bhatia.

Four fact-checks involved claims of Bangladeshi or Pakistani flags at opposition protests, two on an ad shoot in Mumbai passed off as Bangladeshis protesting and two on AI images of the London anti-SIR protest led by Dhruv Rathee, shared by BJP Delhi. Old clips of Hindutva groups were also captioned to urge violence against CJP supporters, the only call for violence among the period's fact-checks.

India-China claims in Arunachal Pradesh

BOOM fact-checked 6 claims about the India-China face-offs in Arunachal Pradesh in the last week of October, 4 of them involving AI. A deepfake of Army officer Girish Kalia, built on 2023 footage of him speaking about infiltration in Kupwara, had him "admitting" that China controls parts of Arunachal. A roundup of visuals linked to the clashes included an AI-cloned voice of Lt Gen Manoj Kumar, and a deepfake of ITBP DG Shatrujeet Kapur was shared by a Pakistani X handle. A video of a religious fair in Nepal was also shared as locals driving out Chinese troops.

Nepal floods: old and AI footage in Indian newsrooms

BOOM published 12 fact-checks on false claims about the late-August floods in Nepal, with 6 using old footage and 5 using AI, including CCTV-style flood clips, a bridge collapse and "rescue" videos of a girl pulled from debris. Mainstream outlets carried 5 of the 12. The Times of India, Republic and News18 were among those that presented a video of separated siblings in Vietnam as Nepal, and Zee News ran a man-made sculpture as a giant Shaligram stone.

Communal claims

BOOM published 14 fact-checks on communal claims (8.5%), none involving AI, and they had the highest share flagged as potentially harmful of any theme.

In August, 4 fact-checks debunked claims about a communal clash in Nepal, all built on recycled footage from elsewhere, such as an old mosque vandalism video from Delhi. Crime stories were also recast with a communal angle, with police contradicting the claims in cases such as a mobile theft in Gwalior.

Crime cases, BRICS and Independence Day

BOOM published 6 fact-checks on false claims about the Jamui (Bihar) molestation case, half of them involving AI, including an AI photo of the accused posing with Akhilesh Yadav. After a building collapse in Delhi's Satya Niketan, the Indian National Congress's official handle shared AI-made and old videos.

Six fact-checks were on false claims about the BRICS summit, 4 of them involving AI. An AI "selfie" of Modi with world leaders was shared by BJP spokesperson Shazia Ilmi and by the Russian Embassy in South Africa.

Seven fact-checks were on Independence Day claims, including an AI letter claiming India's independence was a 99-year lease, and an unrelated video shared as BJP leaders hoisting the party flag instead of the tricolour, which led to an FIR against Congress's Supriya Shrinate and others.

Newsrooms, Officials And Recycled Media

Mainstream outlets themselves carried the false claim in 32 of 165 fact-checks (19.4%), against 37 of 527 (7.0%) in January–July. News18 and ABP News were named most often, followed by NDTV, Republic and Zee News. ANI, NDTV, Times Now and ABP News reported that Akhilesh Yadav had swapped his red cap for a saffron turban, but the turban was red. ABP News apologised and removed a video that falsely linked singer Neha Singh Rathore to a sex racket, and Zee News deleted its posts after Bangladesh's Attorney General denied that the country had banned Jamaat-e-Islami.

Varanasi police's official handle described a real video of policemen filming dancers as AI and announced an FIR against those circulating it, but BOOM verified that the video was real. With the two chatbot cases above, BOOM has documented six cases in 2026 in which an official account or a chatbot called real media fake or old, three of them in these three months.

65 of the 165 fact-checks (39.4%) used media older than the event, against 47.1% in January–July, and 30 (18.2%) were smear campaigns, against 10.8%.