Why the wakeup call should have come long before the 25M devices ad fraud exposé

AI has industrialised ad fraud, creating a harsh reminder that just because something "looks legit", doesn't mean it actually is. While schemes such as Genisys, uncovered by Integral Ad Science (IAS), infected over 25 million devices worldwide, the industry is approaching the threat with caution rather than alarm.

Genisys involved 115 mobile apps and nearly 500 AI-generated websites, designed to mimic blogs, news outlets, and informational platforms. Devices were hijacked to run hidden activity in the background, generating traffic that appeared real but served no legitimate purpose. The scale and sophistication of the operation highlight just how easily AI can scale fraudulent activity across digital ecosystems.

The fallout has prompted some brands to rethink media evaluation, especially as trust becomes a valuable currency.

Don't miss: 101 on synthetic data and why it could be big in advertising

For brands such as Singtel, for example, the trust conversation with partners have been a longstanding one which forces its agencies to go beyond reach as a metric of measurement.

"We're asking our agencies and partners tougher questions about the provenance of their data and the authenticity of the inventory they buy on our behalf," said Lynette Poh, head of brand engagement and loyalty, Singtel. She added that the brand has moved away from top-of-funnel vanity metrics that bots can easily trigger, and instead focuses on predictive models for engagement - looking at how media spends drive actual behaviour on its platforms.

"A click is meaningless if it doesn't come from a real customer [...] It’s a priority for us to ensure that our marketing spend isn't just efficient, but ethical and effective," she said. Actively, Singtel has been prioritising curated, premium ecosystems, seeking smaller, verified audiences over massive, potentially synthetic reach. Poh described it as moving towards a "quality first framework" where the telco giant is investing in environments that respect the consumer's attention and the brand's investment.

As AI evolves, the industry also needs to strengthen how it safeguards trust across the overall digital ecosystem, said Lynn M Chong, head of marketing, Samsung Electronics Singapore. Synthetic inventory and AI-driven ad fraud show how quickly bad actors can exploit new technologies, and for marketers, this reinforces the importance of treating media governance and verification with the same seriousness as cybersecurity, Chong explained.

"The key is not simply concentrating spend, but ensuring campaigns run in verified environments where impressions represent real audiences. We place strong emphasis on independent verification, brand safety standards and fraud monitoring tools," she said. Beyond impressions, Samsung Electronics also prioritises signals tied to meaningful engagement - whether that’s consideration, store visits or real conversions.

A move to premium ecosystems?

With rich engagement set to grow in importance, marketers MARKETING-INTERACTIVE spoke to have varied thoughts on whether the current report on AI-driven ad fraud would prompt major budget shifts.

Cyril Prasanna Kumar, regional digital marketing head, ASEAN, Reckitt, foresees budget shifting, but only for a bit, and for a short term only. "When stuff like this pops up, you naturally lean towards platforms that feel more controlled, but I wouldn't swing 100%," he said, adding that open web still has value.

The bigger question, said Jayss Rajoo, director of marketing and food innovation, is not whether or where the budget would shift to, but rather around the quality of supply and transparency of the path to inventory, as well as my marketing and business objectives.

“Walled gardens naturally offer more controlled environments, but they also come with their own limitations around transparency and measurement,” she said. At the same time, the open web still plays an important role in reach and discovery. What this really reinforces, she added, is the need to be more deliberate about where scale comes from - curated marketplaces, trusted publisher relationships, or tighter supply path optimisation within programmatic buying.

“The objective is not to simply to move budget into walled gardens, but to ensure that scale is coming from environments we understand and trust, rather than from opaque inventory that looks efficient but may not reflect real human attention,” she said.

What is also key is for marketers to have a good understanding of their funnel. "If you don’t, that’s where you’ll get caught in these kind of frauds," said, fractional CMO and Quno AI advisor Delbert Ty. For example, if a brand is looking at a single metric, it should ideally look at the bottom of the funnel metric and work its way upwards.

"If you’re looking to drive purchase and you know your funnel includes some kind of login and people clicking your website and so on, then you look at areas such as CPM, click through rates bounce rates, and if any of those matrix look weird, and when I say weird, it could look really good or really bad then that’s worth digging into," he added.

The need for the industry to align

With the threat of AI-driven fraud reshaping where and how media budgets are allocated, there is a real need for the industry to come together. "As AI capabilities expand, marketers are increasingly focused on protecting the integrity of digital advertising. This requires closer collaboration across platforms, technology providers and brands to ensure transparency, accountability and trusted AI ecosystems moving forward," said Samsung Electronics' Chong.

Kumar described the challenge as “the scary part” of AI-driven media where basic metrics can no longer be trusted. Instead, he looks at deeper engagement signals such as time on site, actual conversions, repeat behaviour, and alignment with first-party data.

"I do think people are starting to worry about AI-generated junk traffic. It feels like one of those things that’s quietly been building… and now stories like this make it more real. While I wouldn’t say we are in a full panic mode, there definitely more awareness than before," he added.

Agreeing with him Rajoo said the conversations around AI fraud are starting, though it is still relatively early in many organisations. " What we do see internally is less about panic and more about raising awareness — making sure teams understand where inventory is coming from, how verification works, and how to interpret media metrics more critically," she said.

Rajoo also shared that Pizza Hut is now actively being more thoughtful about how AI is used in media buying workflows. "Automation can bring efficiency, but it is still important to ensure human oversight and proper checks and balances, especially when large volumes of media decisions are being made by algorithms," she said, adding:

Ultimately, it will push the industry to focus more on trusted supply, stronger verification, and outcomes that reflect real human behaviour rather than just media activity.

Be part of #Content360 Singapore, 22–23 April 2026, where creativity and culture collide. Explore how AI-driven storytelling is shaping the future of content, gain practical insights, discover new tactics, and learn how the best in Asia are creating campaigns that truly resonate. 

Related articles:
Synthetic data for dummies: 101 on what it is and why marketers should care
Indonesia emerges as safe haven for advertisers amid global ad fraud surge, IAS says
Australia leads in video viewability, but brand safety and ad fraud remain key concerns