Ad Tech Briefing: Publishers are turning to AI-powered mathmen, but can it trump political machinations?

Inevitably, this caution leads to a “set-and-forget” mindset that leaves money on the table, often directed into the pockets of Big Tech rather than legacy media owners, contributing to the subsequent decline of the news publishing sector.

In practice, it has often been enforced through blunt technical defaults: keyword blocklists, conservative suitability thresholds, and verification rules that treat “news” as a proxy for danger. 

The result has been a structural shift of budgets toward Walled Gardens, where advertisers continue to appear alongside news-adjacent content, but under different measurement regimes and with limited transparency – just think of how news content is often the topic of trending topics on social networks.

That dynamic is now being questioned — less on ideological grounds than on economic ones.

A growing number of publishers and ad tech challengers argue that news has been systematically over-blocked by legacy brand safety systems, suppressing supply, inflating prices elsewhere and depriving advertisers of engaged audiences. 

Recent moves by Hearst, which has adopted contextual measurement from Mobian across its news and television properties, reflect this shift. Hearst claims that a substantial share of its inventory has been misclassified as unsafe by legacy verification techniques, despite analysis showing that the overwhelming majority of impressions run in neutral or positive contexts.

However, challengers are now positioning their products less as moral arguments for news and more as technical corrections, an approach established with market hegemons, i.e., DoubleVerify and Integral Ad Science. 

These outfits, such as Mantis and Mobian, contend that context-aware, sentiment-driven measurement can distinguish between genuinely risky content and routine reporting — and, in doing so, unlock inventory that has been excluded by default. Importantly, the promise is not merely better monetization for publishers, but also lower CPMs and improved reach for advertisers competing over a narrower pool of “safe” impressions.

Incumbents are also adjusting. Both DoubleVerify and Integral Ad Science have softened their public posture toward news, emphasizing suitability over safety and encouraging buyers to use their tools with greater precision. That evolution reflects a recognition that rigid, one-size-fits-all controls have created inefficiencies — and that advertisers are increasingly willing to test alternatives, provided the risk is measurable and defensible.

New math, same market forces

This recalibration is happening alongside a broader rethink of measurement itself. At last week’s IAB Annual Leadership Meeting, the IAB announced Project Eidos, a multi-year initiative to modernize ad measurement across channels. The project, backed by major brands, agencies, platforms, and measurement vendors — including DoubleVerify and IAS — aims to replace fragmented, channel-specific models with interoperable frameworks built on shared definitions and standardized inputs.

The timing matters. According to the IAB’s State of Data 2026 report, a majority of buy-side practitioners believe that current advanced measurement tools fall short in terms of trust, rigor and timeliness. At the same time, many see AI-driven improvements as capable of unlocking tens of billions of dollars in media investment within the next two years. In that context, over-blocked news inventory looks less like a values stand and more like a measurement failure.

None of this guarantees a wholesale reallocation of budgets away from platforms such as Meta, Google, or TikTok. Matt Prohaska, CEO of Prohaska Consulting, said one of the most meaningful signals to emerge from the IAB’s Annual Leadership Meeting was a growing willingness among advertisers to judge news on ROI rather than sentiment alone, particularly as measurement standards begin to move beyond last-click attribution. He argued that as long as buyers rely on bottom-funnel scoring, news — like other traditional publishers — will remain structurally disadvantaged relative to Google, Meta and Amazon, where performance is assessed under very different rules.

Set and forget

Prohaska, whose firm is behind the ProNews Collective, said marketers are increasingly paying attention to the “messy middle” between awareness and conversion, which is beginning to rebalance how news is evaluated. “As long as buyers continue to focus on last click, last touch, news is always going to have a disadvantage — just like every other so-called traditional publisher — against Google, Meta and Amazon when the score is kept that way,” he said, adding that broader adoption of advanced measurement by brands such as Mars Wrigley and Unilever, should mean agencies start giving news more of the credit for ROAS, than it currently gets.

However, political pressure breeds agency risk-aversion, and “set-and-forget” brand-safety policies remain powerful forces. But the argument is shifting. Instead of asking advertisers to support news because it is civic-minded or principled, publishers and their partners are asking a narrower question: are buyers leaving performance, efficiency and reach on the table because of outdated measurement assumptions?

If improved, empirically grounded measurement can answer that convincingly, it may succeed where moral suasion has struggled — not by challenging advertisers’ politics, but by challenging their math.

Ghost at the feast

However, for all the talk of change, there are two ghosts at the feast, with separate developments away from proceedings at ALM underscoring the scale of the challenge facing publishers. 

Recent filings by a former WPP executive, alleging mistreatment over refusal to engage in the R-word (yes, rebates), demonstrate how there’s a different kind of math at play. 

Meanwhile, OpenAI’s ChatGPT ads beta is proving to be one of the hottest media rollouts in recent years – so much so that it has seemingly foregone an extended agency roadshow for now. 

Since the beginning of 2026, Digiday contacted more than 23 agencies, with most reporting that their knowledge of the fledgling ChatGPT ads business came from third parties rather than OpenAI directly. 

Hence, the media ambitions of LLMs seem to challenge the market hegemony of even Silicon Valley’s incumbents, a development that will make the race for brand spend a three-way contest.  

Numbers to know

Findings of a recent survey of U.S. agency and brand media buyers, by the IAB.

• 96%: of buyers are aware of agentic AI for ad buying and campaign execution.  
• 66% report an increased focus on agentic AI for media buying and campaign execution in 2026.  
• 72%: prioritize cross-platform measurement as a key focus area, up from 64% the prior year.  
• 57%: plan to put increased focus on creator and influencer ads or partnerships in 2026.  

What we’ve covered

Lego is building out an in-house programmatic team

Lego is hiring programmatic experts, building on roles it first recruited for in 2024, although it maintains it is still working with its agency partner Publicis Media. 

A running list of publisher lawsuits targeting Google’s ad tech practices

Digiday has compiled a running list of publishers’ lawsuits against Google for its ad tech practices.

What we’ve heard

“They’re ignoring the “Yahoo Lesson.” Decades ago, portals like Yahoo sold search as a category by bundling inventory and selling top-of-page banners at ~$15 CPMs. The yield (eCPM) was terrible.” 

— TvScientific (now a Pinterest entity) chief executive Jason Fairchild notes how “OpenAI’s move into advertising feels like 1998 all over again… and not in a good way” on his LinkedIn profile.  

What we’re reading

Blackstone-backed Liftoff postpones IPO on software rout

Liftoff Mobile shelved its initial public offering after tech stocks declined amid concerns about the impact of artificial intelligence on operating companies, a sign that IPOs will remain a thing of the past for the ad tech sector.

A former WPP exec is suing the ad agency giant, claiming he was fired after flagging an alleged kickback operation

A former executive at WPP is suing the advertising agency giant, alleging that the company retaliated against him and terminated him after he raised concerns that the group’s media investment division was allegedly running an improper kickback operation.

OpenAI confirms $200,000 minimum commitment for ChatGPT ads

OpenAI is asking select advertisers to commit at least $200,000 as part of a small, tightly controlled beta advertising program.

OpenX appoints Matt Sattel as CEO
Matt Sattel steps into the role as chief executive at the supply-side platform, after the tragic death of OpenX’s earlier CEO, John Gentry, last month.