This Ad Tech Briefing covers the latest in ad tech and platforms for Digiday+ members and is distributed over email every Tuesday at 10 a.m. ET. More from the series →
During the inaugural annual general meeting for The Agentic Advertising Organization, dubbed The Foundry, last week, attendees debated an awkward talking point for the advertising industry’s AI ambitions.
For all the enthusiasm around “agentic” systems, which can plan, optimize, and measure the effectiveness of media campaigns with minimal human intervention, i.e., realize the headcount savings currently wreaking havoc on adland’s workforce, one topic that will thrill accountants less so keeps arising: accountability.
That was the recurring subtext across presentations by executives, academics, former leaders of industry trade bodies, and contemporary practitioners. Namely, if AI agents are increasingly making marketing decisions, who ultimately owns the consequences when those systems fail, bias outcomes, optimize toward the wrong goals, or simply produce mediocre work at scale?
The anxiety was not limited to a handful of directors; the topic was widely debated by the majority of attendees during workshop sessions at the May 6 event. Speakers repeatedly referenced the industry’s tendency to automate flawed processes rather than redesign them, with former IAB U.S. CEO Randall Rothenberg presenting findings that 95% of generative AI pilots fail to deliver measurable profit-and-loss impact, and another forecast that 40% of agentic AI projects will fail by 2027.
Optimizing toward average
Backing up this attitude, fellow presenter Colleen DeCourcy, CMO, Sonos, issued a warning about optimization culture. “The future of creativity will not be automated,” she said, taking aim at some of the shortcomings of programmatic advertising. “Programmatic is really good at attributing value to people who were already on their way to you,” she said, arguing the industry had spent years prioritizing measurable efficiency over originality and persuasion.
Furthermore, DeCourcy described AI-generated creative as optimized toward adequacy, i.e., technically functional, brand-safe, and scalable, but lacking the unpredictability or emotional resonance that builds durable brands. “It can harvest a field,” she added. “It cannot plant one.”
Not to be deterred, few speakers argued against automation outright, with much of the discussion focusing on governance frameworks designed to constrain increasingly autonomous systems before they become operationally embedded.
That was especially evident during presentations, with the prevailing sentiment during The Foundry’s workshop sessions, conducted under the Chatham House Rule, summarized by one group spokesperson, “An agent shouldn’t evaluate its own work.” A further recurring principle voiced during the sessions was that efforts needed to be made to prevent closed-loop optimization systems from effectively grading themselves without external oversight.
That includes establishing guardrails, measurable goals, audit guidelines, and legal accountability structures for agentic systems operating across media and commerce environments, according to conference attendee Ruben Schruers, CEO, Ebiquity.
“Let’s say, a buyer or seller agent makes a mistake and does something it’s not supposed to do,” he told Digiday. “Is it ChatGPT or Claude [who is legally responsible]? Or is it the licensing or maintaining the user-agent, or the exchange system on which the agents engage with each other?”
The emphasis on governance also reflected broader industry concerns about the concentration of power among large technology platforms, with certain presenters outlining a more glass-half-full outlook to the agentic era.
During his presentation, Brian O’Kelley, Scope3 CEO and founder of AAO, argued that AI agents could eventually reduce the ad industry’s dependence on dominant platforms by enabling more interoperable infrastructure across the open web. “The industry needs to reduce reliance on Big Tech,” he said, adding that agentic systems could make that possible “at scale.”
Whether the industry can actually operationalize that vision remains uncertain. Several speakers acknowledged that marketers themselves often remain organizationally fragmented, with creative, media, measurement, and commerce functions still operating independently despite AI systems increasingly connecting them.
So, with the rapid advances of such technology, the key question — and one which overshadowed AAO’s event — is whether the industry has developed the institutional accountability required to manage it?
What we’ve heard
“There was a rumour [sic] last year that you wouldn’t be able to enter the hotel or lobby bar without a conference badge. That was completely false. There were about 20 guests not affiliated with POSSIBLE. I spoke to one, and they were not happy as they weren’t told about the conference.”
– Those attending last month’s POSSIBLE conference in Miami, FL, heard as many rumors as they did sales pitches, but the extent of its expansion is still subject to speculation.
Numbers to know
$1.6 billion: The amount Yahoo raised from a loan and bond deal to refinance debt tied to its acquisition by Apollo Global Management.
$1.8 billion: AppLovin’s Q1 revenue, up 59% year-over-year.
$466.4 million: Taboola’s Q1 revenue, up 9.1% year-over-year.
$14.8 million: Cindy Rose’s potential annual remuneration approved at WPP’s investor day last week.
What we’ve covered
The Trade Desk’s $689M revenue beat is blunted by the departure of CSO Jacobson to OpenAI
The Trade Desk’s Q1 revenues of $689 million may suggest holding-company spats are a mere bump in the road, but confirming the exit of CSO Samantha Jacobson hours before its market disclosure shows complications can come from all directions.
The upfront is no longer just a shop window for content Amazon owns, it’s now one for its ad tech too.
What we’re reading
Magnite says it’s not afraid of the agentic era – or of Google
Magnite posted Q1 revenues of $164.4 million, primarily driven by gains in its CTV business, but its traditional display ads business saw a 5% annual decline during the period.
Despite enthusiasm over its ChatGPT tie-up, Criteo’s shares slide on downgraded revenue forecast
The buzz around Criteo’s groundbreaking partnership with OpenAI was palpable during the ad tech company’s first-quarter earnings call earlier today, but an earlier-tipped lag in retail media spend, plus slowed performance media spend, impacted its revenue during the quarter.
Rising Omnicom star painter out: Gambino, McCord succeed Omni, Commerce respectively
Omnicom announced long-time Flywheel Digital executive Christine Gambino has been named CEO of its homegrown platform Omni, succeeding Duncan Painter, who is leaving Omnicom, while long-time Flywheel CEO Alex McCord will now lead Omnicom Commerce.
PubMatic’s agentic AI is going beyond direct deals
PubMatic has run more than 30 fully autonomous, end-to-end agentic campaigns through the SSP’s AgenticOS platform, CEO Rajeev Goel told investors during the company’s Q1 earnings call on Thursday.
More in Media Buying
Amazon positions live sports portfolio as leverage to pry open upfront dollars
Amazon touted its combination of commerce targeting data and NBA and NFL inventory as a solution to marketer wariness during upfront negotiations.
Why Duluth trusts AI agents with bidding, but not brand storytelling
This article is part of a series covering our Programmatic Marketing Summit. More from the series → There seems to be two schools of thought when it comes to agentic media buying. There are those who want to be at the so-called cutting edge, embedding AI agents in things like the process of buying ads […]
Media Buying Briefing: Takeshi Sano is making the rounds for Dentsu — is it enough?
Dentsu just retained Heineken’s global business, and apparently the new CEO was involved. Should he take the heat for losing Microsoft? Most say no.