This festive season, AI-powered ad fraud is set to hit harder than ever

New Delhi: India’s most high-stakes marketing season is around the corner, spanning Ganesh Chaturthi, Onam, Navratri, Diwali, and Christmas. 

Often dubbed the “Olympics of Advertising,” brands go all out to lure consumers. But beneath the festive glitter lies a growing menace: AI-powered ad fraud. 

With a 10–12% rise in festive season adex expected this year, industry leaders are also raising concerns about the growing threat of AI-driven ad fraud. 

“What’s really at stake isn’t just wasted media budgets; it’s the illusion of success during festive seasons,” said Vivek Kumar, Chief Strategy Officer at DViO Digital.

Citing Juniper Research data, Kumar said that global digital ad fraud is projected to exceed $100 billion annually by 2025.

In India, WPP Media's This Year, Next Year report estimates that digital ad spend will cross Rs 82,000 crore in 2025, with nearly Rs 30,000 crore spent during the festive quarter alone, primarily by sectors such as e-commerce, FMCG, auto, and consumer tech, he added.

"Even a conservative 10% loss during this window puts Rs 3,000 crore at risk—lost to bots, fake leads, invalid traffic, and non-human engagement," Kumar emphasised.

Bharatesh Salian, President – Digital at Wondrlab, recounted what happened to a mid-sized electronics client during last Diwali:

“Every brand is on steroids during the festive quarter, each campaign trying to be the next Shah Rukh Khan of advertising. Amid glittering creatives and ambitious KPIs, a quiet villain sneaked in: ad fraud.”

Despite a solid product, sound strategy, and a campaign mix spanning programmatic, reels, and influencers, the client saw a huge disconnect. Clicks surged. Installs spiked. But sales didn’t budge.

“It turned out bots were filling forms, clicking links, even installing the app. One fake shopper completed the entire onboarding journey in 1.7 seconds. Thankfully, fraud detection tools running in the background flagged anomalies, paused rogue sources, and saved nearly Rs 24 lakh in ad spend,” Salian said.

While AI has helped marketers fine-tune targeting and optimise outcomes, it's also equipping fraudsters with smarter weapons. From bots that mimic human behaviour to click farms using generative content, AI has industrialised ad fraud at scale.

“If the data seems too good to be true, stop and question it. Without anomaly detection, everything looks like it’s working, but behind the curtain, something else is happening,” said DViO’s Kumar.

“Today’s bots scroll like users, tap like users, fill lead forms like users. They behave like your dream customer, just without the wallet. And during the festive rush, when scrutiny is low and urgency is high, this fraud thrives.”

McAfee reported that 45% of Indian shoppers encountered deepfake scams during the last festive season. mFilterIt estimates that 23% of programmatic traffic is junk. However, brands using proper detection tools saw fraud rates drop by up to 36%.

“AI gives us sentinels: machine learning tools that don’t just detect fraud, they predict it,” said Salian.

“Anything driven by algorithms can be gamed by another algorithm. Platforms like mFilterIt and DoubleVerify are now like X-ray machines for ad campaigns. You may be chasing impressions, but they’ll tell you who’s wearing a mask.”

Yet not all experts are confident that current tools are enough.

Rajiv Dhingra, Founder and CEO of ReBid, cautioned that most tools in the market today are “rule-based or reactive,” flagging fraud only after damage is done.

“Bots are no longer basic scripts. They’re trained models capable of bypassing traditional fraud filters. GenAI is even being used to generate synthetic influencer content and fake user reviews that fool both platforms and consumers,” he said.

Deepit Purkayastha, Co‑founder of Inshorts, added that fraudsters are also leveraging AI to drastically reduce the cost of deploying these fake environments, enabling mass exploitation of ad supply chains.

And evolution isn’t just technical, it’s creative. As Dhiraj Gupta, CTO of mFilterIt, highlighted, ‘We’re seeing deepfake creatives used in spoofed influencer campaigns and smart bots orchestrating affiliate fraud using ML‑optimized spoofing techniques.’ These aren’t static fraud schemes, they’re dynamic, adaptive systems that continue to evolve and evade.”

During Diwali 2024, some large advertisers saw 18–22% of their programmatic traffic deemed invalid after deeper audits, representing crores of wasted impressions.

“AI-powered fraud demands AI-powered defences that continuously learn, adapt, and act in real time,” Dhingra added.

Moneka Khurana, Country Head, MMA Global India, also shared a telling example: “A large D2C client saw a 25% dip in measurable ROI last Diwali. Post-campaign audit revealed that 70% of suspicious activity came from bot farms in mobile web placements.”

Saurabh Khattar, Country Manager, India at Integral Ad Science (IAS), echoed these concerns. “Traditional fraud detection tools, which rely on static lists or rules, are no longer sufficient. Fraudsters now use AI to build fake websites, user agents, and synthetic content that looks legitimate but exists only to siphon programmatic revenue.”

He noted that during IPL 2025, IAS observed a “93% spike in invalid traffic” between February and March, highlighting how fraud scales around marquee events.

“We need adaptive, multi-layered protection that uses behavioural, deterministic, and contextual models to learn from new threats, like falsified user agents or made-for-advertising (MFA) sites,” Khattar said.

Akshay Chaturvedi, Chief Business Officer at VDO.AI, stressed that no single channel is immune.

“Programmatic advertising faces structural challenges due to its automated nature, with fraud rates touching 46.9% in certain networks. Social media platforms like Meta and Google invest heavily in safeguards, but fraud still slips through. Influencer marketing brings its own risks. While 58.5% of mega-influencers may engage in metric enhancement, nano and micro-influencers often deliver more authentic engagement,” he said.

Khurana issued a broader warning: “The festive season is India’s Super Bowl of advertising. Budgets double or triple. But as brands chase reach, strategic precision often takes a back seat. The biggest risk is not just fraud; it’s irrelevance.”

Chaturvedi laid out the economic impact with hard numbers. “Industry data shows that up to 25% of paid traffic is fraudulent. In December 2024, fake clicks accounted for 13% of ad traffic. With 30% of annual media budgets spent in Q4, that translates to over $600 million in lost global ad spend during the holiday season alone.”

Fraud rates spike by 35–40% during periods like Diwali, Dussehra, Black Friday and year-end sales. Impressions surge by 50%, click-through rates double, and malicious activity gets masked by volume.

“Small businesses are hit hardest, losing up to 30% of festive ad budgets to fake traffic,” he added.

As AI-powered fraud grows more sophisticated, experts say standard platform filters and post-campaign audits are no longer enough. A proactive, full-funnel strategy is now essential, from planning to real-time monitoring.

“We need to stop thinking of fraud as backend hygiene and start treating it as a frontline business threat,” said Kumar.

“Audit your funnel end-to-end. Don’t rely blindly on platform-level filters. Google and Meta protect their platforms, not your ROI. Move beyond last-click attribution. Bots love that loophole. Be cautious with high-incentive campaigns; high payouts attract high fraud. Build anomaly detection into your workflow. And above all, define success beyond surface metrics.”

Khurana wrapped up with a clear directive for decision-makers: “Brands must treat ad fraud not as a vendor issue, but a boardroom priority. As festive budgets spike two to three times, so must our vigilance. AI is the new threat, but it’s also our best defence.”

“Building a real-time fraud detection ecosystem should not be seen as a cost,” she added. “It’s an ROI enabler.”