CTV ad fraud surges 140 per cent as AI-powered schemes spread globally

CTV ad fraud schemes and variants surged 140 per cent in the March quarter compared with the same period a year earlier, according to new research from media verification provider DoubleVerify, as bad actors increasingly deploy artificial intelligence to exploit gaps in the fast-growing connected television advertising market.

Published 7 May and covered by Adweek, the study identified more than 50 distinct bot attacks and variants targeting CTV inventory during 2025, alongside a tenfold increase in fraudulent CTV apps compared with the year prior. Unprotected CTV campaigns carry a fraud rate of roughly 9 per cent — translating to losses of about $US1.8 million ($A2.7 million) per billion impressions — against less than 1 per cent for campaigns that use independent verification tools.

The surge comes as connected TV — advertising delivered through internet-connected television sets, streaming devices, and smart TV apps — has become one of the fastest-growing channels in digital advertising. Unlike traditional linear TV, CTV inventory is bought and sold programmatically across a web of exchanges, supply-side platforms, and demand-side platforms. But each link in that chain adds a layer of complexity that fraudsters can exploit. And the money flowing through it is climbing fast. Global CTV ad spend is projected to exceed $US40 billion in 2026, according to industry estimates, making the channel an increasingly attractive target for sophisticated fraud operations.

“CTV is attracting premium spend and bad actors right along with it,” said Gilit Saporta, vice president of DoubleVerify’s Fraud Lab. “Our research shows fraudsters are quick to exploit inefficiencies in the market, using AI and limited transparency to siphon value from advertisers, with tactics that vary by market.”

How fraud breaks down by region

Across North America, bot fraud accounts for 82 per cent of fraudulent activity, the study found. In Asia-Pacific, data centre traffic — where fraudsters route impressions through server farms to mimic legitimate viewers — represents 98 per cent of detected fraud, the highest share of any region studied. Europe, the Middle East and Africa recorded 66 per cent data centre traffic fraud, while Latin America registered 91 per cent.

Australian advertisers, in particular, should pay attention to that 98 per cent APAC figure. The country’s programmatic CTV spend flows through many of the same exchanges and supply paths that carry tainted inventory across the region, SecurityBrief Australia reported, even if DoubleVerify’s study did not isolate local market data. But the Interactive Advertising Bureau’s Australian arm has been working with members on updated measurement guidelines for CTV. Mandatory verification requirements, however, remain limited compared with markets such as the United Kingdom, where the Joint Industry Committee for Web Standards has pushed harder for anti-fraud tool adoption.

Direct deals are not a shield

The findings challenge a widely held assumption in the advertising industry: that direct deals between buyers and CTV publishers are inherently safe from fraud. In one consumer healthcare campaign studied by DoubleVerify, 34 per cent of impressions served through a direct deal went to bots. A separate consumer packaged goods campaign recorded 25 per cent bot traffic under a similar arrangement.

“There’s a perception that direct deals in CTV are fraud-free, but that’s not the case — fraud always finds a way,” Saporta said. “It can exist anywhere inventory is bought and sold. Without independent verification and proactive protections, advertisers risk paying premium prices for impressions that deliver no real value.”

They undercut what has been one of the industry’s default risk-management strategies. Programmatic buyers who assumed they had sidestepped the fraud problem by negotiating directly with publishers are learning that the bots do not respect the contract type.

AI on both sides of the ledger

And DoubleVerify’s researchers noted that AI tools are enabling fraudsters to generate convincing app listings, rotate device fingerprints to evade blocklists, and simulate human-like viewing patterns that basic verification systems struggle to catch. That 140 per cent increase in scheme variants reflects an attacker operation iterating faster than most advertisers’ defences can keep pace with.

Yet AI is also powering the defence. DoubleVerify, listed on the New York Stock Exchange with a market capitalisation of roughly $US5 billion, has been expanding its CTV verification capabilities as the fraud landscape evolves. The company’s Fraud Lab — the unit Saporta leads — uses a combination of machine learning, traffic pattern analysis, and manual investigation to identify and block fraudulent activity across connected TV, mobile, and desktop environments.

Advertisers who have not yet adopted independent verification face unforgiving arithmetic. At 9 per cent fraud on unprotected inventory and $US1.8 million in wasted spend per billion impressions, every campaign running without verification is handing a measurable cut to the bots. For Saporta, the prescription is blunt: verify everything, assume nothing, and stop treating direct deals as a fraud-free shortcut.