Every day, a small team at India Today sifts through a flood of doctored images, out-of-context videos, and AI-generated fabrications — an altered archival photo of a politician here, a deadly flood clip stripped of its real context there, a full-blown deepfake of a business tycoon somewhere else. The six-person fact-checking team works to debunk a slew of false and misleading images circulating across social media every day. It’s a scene that repeats itself, in different forms, across dozens of Indian newsrooms every single day. What has changed in the last couple of years is not the volume of misinformation — that has only grown — but the sophistication of the tools newsrooms now have to fight it.
From Manual Debunking to AI-Assisted Verification
For years, fact-checking in India relied heavily on manual labour: reverse image searches, painstaking frame-by-frame video analysis, and calls to sources to confirm whether an event actually happened the way it was being described online. That process still forms the backbone of verification work, but artificial intelligence is increasingly being layered on top of it to speed things up.
Backstory, an experimental tool from Google’s AI research lab DeepMind, is now helping India Today’s team sort through a stream of scams and disinformation by automatically checking whether an image is AI-generated, detecting signs of manipulation, and tracing where else the image has appeared online. The tool is still limited in availability — it remains part of Google’s Trusted Testers Program, with India Today’s fact-checkers among thousands of journalists, OSINT experts, and researchers piloting it and feeding back improvements — but it signals a broader shift: global tech companies are now building verification tools with journalists in the room, rather than after the fact.
Homegrown Fact-Checkers Lead the Charge
India’s fact-checking ecosystem didn’t wait for Big Tech to hand it solutions. Organisations like BOOM and Alt News have spent close to a decade building their own verification muscle, and more recently, their own technology stacks. BOOM, an IFCN-certified fact-checker operating across English, Hindi, and Bengali, was the first Indian fact-checker to launch a WhatsApp helpline where readers can send in posts and forwards for verification, a line that now receives thousands of queries every week.
That WhatsApp-first approach matters enormously in a country where forwarded messages, not links or articles, are often how misinformation actually travels. Alt News similarly relies on its WhatsApp tipline, alongside social media monitoring, to catch hate speech and circulating disinformation that readers flag directly. These tiplines function as an early-warning system, surfacing the claims that are gaining traction before they spiral into full-blown viral falsehoods.
Detecting Deepfakes at Scale
Perhaps the starkest tech shift has come in response to deepfakes. As AI-generated audio and video have become cheap and easy to produce, newsrooms and platforms alike have had to build specialised detection infrastructure. WhatsApp partnered with the Misinformation Combat Alliance to launch a dedicated fact-checking helpline aimed squarely at AI-generated misinformation, built around a four-pillar approach of detection, prevention, reporting, and awareness. Users can forward suspicious audio notes and videos and receive an assessment of whether the content shows signs of AI generation.
This kind of infrastructure was tested heavily during recent election cycles. The Deepfakes Analysis Unit’s WhatsApp tipline, launched ahead of the 2024 Lok Sabha elections, received hundreds of user messages asking for help determining whether videos were real, with most flagged content turning out to be doctored clips where an altered voice track had been lip-synced onto the original footage using generative AI. Fact-checkers have also learned to combine several tools rather than trust any single detector. BOOM’s own verification toolkit for journalists points to a stack of resources — including AI-detection tools, deepfake analysers, and voice-clone checkers — while cautioning that these indicators are helpful but not foolproof, and must always be cross-checked against multiple sources.
Why This Moment Is Different
What makes the current wave of adoption notable is the sheer scale of AI-generated content fact-checkers are now dealing with. Data from fact-checking organisations elsewhere gives a sense of the trajectory: globally, the share of fact-checked claims involving AI-generated content has more than doubled year over year, driven largely by fabricated visuals spreading widely on video platforms. Indian newsrooms have lived this acutely — during the India-Pakistan conflict of May 2025, disinformation surged almost instantly, with old videos from unrelated explosions repurposed and framed as fresh evidence of attacks, sometimes claimed by both sides as proof of the other’s aggression.
That environment has pushed fact-checking from a niche editorial function into something closer to newsroom infrastructure. Industry researchers now frame this as a matter of survival rather than differentiation — a recent industry study argues that AI is merely an enabler, and that the real competitive advantage for Indian newsrooms in the years ahead will not be speed of publishing but whether audiences actually believe what they read, see, and hear.
The Road Ahead
None of this suggests the problem is close to being solved. Detection tools still lag behind the pace of AI-generation tools, deepfakes are getting harder to spot with the naked eye, and much of the disinformation that matters most in India travels through closed WhatsApp groups where fact-checkers have limited visibility. Even institutional efforts have run into friction — attempts to set up a government-backed fact-checking unit have faced legal challenges over how “misleading” content should even be defined.
Still, the direction is clear. Indian newsrooms are no longer treating fact-checking as an afterthought bolted onto the end of the reporting process. It’s becoming embedded — in the tools reporters use daily, in the tiplines readers rely on, and in the partnerships being built between newsrooms, platforms, and AI labs. In a media ecosystem where a single doctored clip can spread to millions before anyone gets a chance to correct the record, that shift from reactive debunking to proactive, tech-assisted verification may be the most important change happening in Indian journalism right now.
