Home News ‘We’re experiencing an enormous uplift’: How generative artificial intelligence is fueling the...

‘We’re experiencing an enormous uplift’: How generative artificial intelligence is fueling the next wave in adtech fraud[19459017]

0
‘We’re experiencing an enormous uplift’: How generative artificial intelligence is fueling the next wave in adtech fraud[19459017]

AI could be fueling a new chapter in adtech fraud — in ways that the industry is not ready for.

According to DoubleVerify’s research, generative AI content farms are stealing publisher ads.txt file to hijack ad revenues. Since January, we have been tracking this activity.

The company has been investigating “AI slop networks and sites” – in particular, “Synthetic Echo”which uses generative AI for mass-production of content and spoof files called ads.txt to siphon off revenue from legitimate publishers. These sites are filled with low-quality AI-generated content that is scraped off of other reputable publisher websites like headlines and images, to create a fake domain which looks almost identical. This is done to drive ad revenues.

DV has uncovered URLs like https://nbcsportz.com/espn24.co.ukcbsnews2.com and 247bbcnews.com which have been designed to mislead ad tech vendors and buyers by mimicking reputable publishers by scraping their content. These sites have been running ads for months.

(19459028]Provided By DV

Fraudsters are now getting even more bold in their tactics. They copy and paste entire ads.txt entry onto their pages, making it look like they’re trusted publisher to major ad buying platforms. It has found 100 sites that look like this.

In a seven-page document, the tech company releases its latest findings. It includes tips for publishers on how they can continue to audit their ads.txt file and ensure that only trusted sellers are allowed to sell to them.

This is a familiar tale with a twist. The AI layer may be new, but the underlying problem is the same: fraud in packaging.

Gilit Saporta is a vp at DoubleVerify Fraud Lab who has been tracking fraud for decades. She said that generative AI amplifies the existing issues. “It’s adtech fraud schemes on steroids,” she said. “Gen AI allows you to replicate and amplify a bad website, which has been set up to spread fake news or hate speech, or to just monetize it, whether it is real traffic or bot traffic.”

In short, generative AI is not creating new types of adtech fraud. It’s simply making the old ones easier to scale and execute. “Gen AI makes it so much easier to make a site look plausible, and replicate it to 1,000 or 10,000 websites a day,” Said Saporta. “We are not necessarily seeing a rise in sophistication, but an enormous increase in scale,” she said.

(Provided by DV)

Thomas Lue Lytzen is the director of ad-sales and technology at Danish tabloid Ekstra Bladet. “It is extremely disturbing that premium publishers are once again being abused by fraudsters in the ad-ecosystem,” he said. “It makes it more important for advertisers to work only with whitelists and not let the buying algorithms choose.”

Ads.txt, introduced by the IAB Tech Lab as a possible fix to help buyers verify who is authorized to sell inventory in 2017, was introduced. It’s not a silver bullet, but it has helped publishers cut down on unauthorized resellers in the hopes that more ad spends will go directly to publishers.

Programmatic is a game that has always been played by whack-a mole, and this game is only moving forward, said Hamish Nicklin. He was a vocal critic of adtech malpractices during his tenure at Guardian News and Media.

He said that the latest issue of what DV called ads.txt plagiarization was not an AI problem but an adtech problem. The pipes are still easy to spoof. “If advertisers want to chase down the lowest CPMs on the open web, then they must accept the risk,” he said. He believes the best solution is to have buyers spend their money in curated, safe marketplaces, and to rebuild direct relationships with premium publishers. Ad tech vendors should also be held accountable for who gets into the exchanges.

Many publishers share the same opinion: there will always new problems that plague open auctions that are nearly impossible to solve. “There will be new issues that fraudsters game,” said Pritesh J. Jumani, Globe and Mail’s head of adtech and performance media and revenues. Publishers are taking advantage of the opportunity to promote direct deals. Pritesh emphasized that unlike the murky waters where spoofed inventories and slop websites thrive, working directly through publishers via private auctions or preferred deals, as well as programmatic guaranteed, not only mitigates risks but also provides buyers with greater transparency.

The ad tech industry has been plagued with inefficiencies for years. From opaque vendor fees to fraudulent stock that steals ad dollars from real journalism, the ecosystem is riddled by bad actors.

The rise of AI-generated content farm plagiarizing legitimate publisher’s ads.txt file marks a troubling chapter in the long story of digital advertising fraud.

Danielle Coffey (CEO of News/Media Alliance) and Richard Reeves (Managing Director of the Association of Online Publishers of the UK) both expressed “significant concerns” about the latest DoubleVerify findings. Both stated their intention to closely monitor the issue and work with industry stakeholders, including IAB Labs to address it. IAB Tech Labs did not respond in time to publish this article.

It’s nothing new for the publishers.

It seems to be in most adtech companies’ interest to keep programmatic going as it is, said Alessandro de Zanche. He is an independent media consultant and a former News UK director of Programmatic. “We have a clear problem: the digital advertising sink is overflowing. The industry has been debating which mop to use instead of fixing the faucet.

This is the most open secret in the industry for many years. Zanche said that if we fixed the faucet, the business models for most ad-tech companies would collapse.

He believes that the first step to cleaning up is to create an advertising environment where publishers, advertisers, agencies, and those who represent them are legally identified. “Then, we can talk about standards and continue to cut off dead branches. De Zanche said that we don’t require AI to achieve this, but a clear commitment, starting with SSPs and DSPs.

https://digiday.com/?p=579068

www.aiobserver.co

NO COMMENTS

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Exit mobile version