Anthony Katsur has spent greater than 25 years in digital media, however 2025 stands out as probably the most hectic interval in his profession. Talking on the AdTechGod podcast on December 29, 2025, the IAB Tech Lab CEO addressed the promoting {industry}’s rush towards agentic AI whereas basic issues stay unresolved.

Katsur characterised the present agentic AI discourse as reaching “a fevered pitch” of hype. Conversations with holding corporations, publishers, and advert tech companies constantly contain stakeholders claiming they wish to “do one thing agentic” with out clear definitions of what meaning.

“I wish to do an agentic CTV purchase. Okay, what’s an agentic CTV purchase? I do not know. I hoped you inform me,” Katsur stated, describing typical exchanges he encounters. The South Park underpants gnome episode offers an apt comparability: the first step entails declaring agentic intentions, step two stays clean, and step three guarantees revenue.

Early implementations constructed on proprietary foundations

The agentic AI infrastructure stays extraordinarily younger. Mannequin Context Protocol, one framework enabling agent communication, is roughly one yr previous. Firms creating agentic options have largely constructed them on proprietary code or legacy programming languages together with C++ and Python slightly than rising protocols like ACP or MCP.

“I am not listening to something about, you recognize, oh, we constructed this in ACP or MCP or agent to agent simply because all these issues are so new,” Katsur defined. Some holding corporations have operated agentic approaches for practically two years, however these inner or external-facing options do not essentially depend on standardized protocols.

Normal Motors ran its first synthetic intelligence-generated commercial not too long ago, adopted by Nike and Coca-Cola’s second AI-created vacation spot. Generative AI already disrupts copywriting and artistic manufacturing. Katsur expects AI to impression the ecosystem most instantly via inventive and copywriting, with agentic AI bringing efficiencies to media discovery, negotiation, planning, and shopping for.

The CEO acknowledges agentic AI introduces superior automation via pure language processing. Interfaces for getting and promoting media will regularly evolve, although conventional consumer interfaces will not disappear. Pure language approaches to describing audiences, transacting towards audiences, and organising campaigns characterize the place the expertise finally leads.

Protocols can not remedy misaligned incentives

Provide chain transparency stays a persistent problem that protocols alone cannot address. Katsur pushed again towards assertions that agentic AI will remedy provide chain transparency points.

“Protocols do not remedy for misaligned incentives and unhealthy actors. They do not like that is not what protocols handle,” he acknowledged. The promoting {industry} nonetheless must deal with fundamentals like provide chain transparency and privateness earlier than leaping to new phases of agentic improvement.

Privateness issues persist regardless of widespread {industry} fatigue with the subject. Legal guidelines have handed at an unprecedented fee, evolving quickly whereas creating full fragmentation inside america. The European Union launched new omnibus laws probably stress-free some privateness legal guidelines, significantly round AI. Regulators proceed issuing fines and subpoenaing corporations, with exercise anticipated to extend in 2026.

Guard rails should handle what occurs when an agent for one firm interacting with an agent for an additional firm violates a privateness legislation. The IAB Tech Lab works with User Context Protocol, donated by LiveRamp, which capabilities as an viewers and signaling buying and selling protocol. Privateness guard rails want integration into brokers to stop stepping on the {industry}’s personal ft.

“We do have to construct privateness guard rails round that. We do want to verify TCF, the worldwide privateness protocol, issues like the info deletion request framework are inbuilt to those brokers,” Katsur emphasised. The Transparency and Consent Framework has undergone multiple updates, with version 2.3 addressing vendor disclosure ambiguity.

Measurement challenges require enterprise alignment

Measurement stays unsolved as an ecosystem drawback. Katsur would not imagine any protocol will remedy measurement challenges, characterizing them as enterprise incentive points slightly than technical issues.

“They are often solved technically if we might all simply agree on a typical set of telemetry to create a baseline of a measurement framework for the {industry}. We may in all probability remedy the measurement problem. A protocol is just not going to do this,” he defined.

The {industry} wants foundational work earlier than embracing the agentic section. Agentic AI would not characterize an end-all answer. Speak has emerged about brokers changing real-time bidding, however RTB is not disappearing quickly.

“RTB is extremely environment friendly when used correctly, the power to examine each single media impression and the viewers hooked up to that and decide a good market worth of that between a whole bunch of patrons and sellers, you recognize, billions of transactions occurring per second. That’s an extremely highly effective protocol,” Katsur stated.

Agentic interfaces connecting with RTB grow to be attention-grabbing, however guard rails stay crucial. Katsur sees genuine options pointed at direct media shopping for and promoting, although he views brokers as significantly efficient for prime cardinality issues like discrepancy decision or troubleshooting inventive points throughout related TV companions.

The idea of agentic interfaces facilities on reworking how customers work together with promoting techniques. Relatively than predefined viewers segments, pure language capabilities of huge language fashions may allow patrons to find extra refined audiences.

“I feel the pure language capabilities of an LLM powering an agentic viewers platform in all probability permits patrons to sus out perhaps extra refined audiences. It in all probability permits sellers to curate perhaps extra nuanced audiences for patrons to purchase,” Katsur defined.

Prompts will more and more complement knobs, dials, and sliders slightly than changing them totally. The comparability to self-driving automobiles applies: fingers stay on the wheel often even with autonomous techniques engaged.

Advert operations faces slowdown in junior hiring

Advert operations departments will doubtless expertise the best impression from agentic AI, although not essentially via downsizing. Katsur expects a slowdown in hiring patterns as a result of automation allows doing extra with much less.

“I am already listening to from some publishers, they’ve closed quite a lot of open roles for junior advert operations folks as a result of AI or Agentic options can help you. It is a device. It permits you to do extra with much less,” he acknowledged.

The standard skilled pyramid mannequin—with broad bases of junior expertise regularly transferring up via lifelong apprenticeships—is reworking right into a diamond form. Fewer alternatives will exist at junior ranges, with large tech corporations not hiring as many junior engineers. AI can write unit exams that junior engineers as soon as dealt with. Junior advert operations duties like reconciling discrepancies between campaigns now get automated via brokers that full in seconds what beforehand took people hours, days, or weeks.

Mid-level administration stays crucial however should grow to be extra hands-on as a result of AI allows higher activity administration, mission administration, and standing monitoring. Enterprise improvement, gross sales, account administration, and technical account administration characterize sound roles now and into the longer term. These people-to-people interactions determining collaboration and income optimization can’t be automated.

Some advert operations professionals will evolve into these relationship-focused roles. Understanding how the engine room works helps affect it correctly utilizing brokers to drive outcomes.

AI browsers reshape content material presentation

AI-powered browsers current two basic implications for promoting. First, they perform as trojan horses for continued crawling from residential IP addresses. Second, they essentially change net design and data structure.

“What’s extremely highly effective is that you just and I could possibly be trying on the similar content material and Atlas may lay it out in another way for every of us. So it is virtually like customized curation of content material. The content material would not change. It is simply the way in which it is offered to every of us could possibly be barely or radically completely different,” Katsur defined concerning ChatGPT’s Atlas browser.

This functionality impacts web page structure, advert positioning, presentation, and shopper expertise. The {industry} might head towards a world the place net servers publish markdown information whereas AI browsers lay out content material based mostly on particular person preferences.

Subscribe PPC Land publication ✉️ for comparable tales like this one


Subscribe

Strategic planning requires industry-wide collaboration

The IAB Tech Lab begins strategic planning for annually in June of the prior yr. The 2026 planning began in early June, involving conversations with members, working teams, and inner staff debates about developments shaping the {industry}.

These discussions get offered to the board via one-on-one conferences at the very least twice yearly with every board member, generally thrice. The tech lab board summit brings collectively complete boards the place board member firm staff can take part. A Magnite board seat may deliver a product lead and engineers to have interaction in strategic conversations figuring out closing roadmaps.

“A few of it frankly is a few of it is a bit of thumb within the air like I feel that is the place the ecosystem goes. So there’s a bit of little bit of artwork to it,” Katsur acknowledged. The group has elevated outreach to holding corporations and publishers this yr in comparison with earlier years.

“We have completed a superb job this yr of actually getting extra holdco suggestions and extra direct writer suggestions than I say we have completed in years prior as a result of I feel generally once we roll out our specs the place the rubber hits the highway is in adoption and that’s often resulting from the truth that we didn’t get principal buyin,” he defined.

Advert tech facilitates how patrons and sellers wish to transact, making their enter crucial for efficient requirements improvement.

First tasks will doubtless fail however innovation continues

The agentic AI bubble dialogue raises questions on whether or not it should pop or deflate. Katsur referenced a latest article noting that first agentic AI tasks will doubtless fail, which represents regular innovation patterns.

Thomas Edison did not invent the sunshine bulb however perfected the filament to maintain mild bulbs illuminated longer. Early movers in agentic approaches will see blended success earlier than some implementations take maintain.

“I feel we will see subsequent failures and successes I feel earlier than we stabilize round an method. And I do not know if there’s going to be one method,” Katsur stated. A number of protocols may assist agentic AI slightly than a single standardized technique.

The innovation interval will span two to a few years, following the Gartner hype cycle’s peaks and troughs. The {industry} at present sits on the peak of optimism and can hit troughs of despair, however agentic AI will not disappear.

“I do not assume it is going to essentially pop I feel we’ll see a deflation,” he predicted. The broader AI ecosystem’s final result relies upon extra on monetary administration and funding between main gamers together with OpenAI, Nvidia, Anthropic, and Perplexity. Some bubbles might burst relying on funding success.

The IAB Tech Lab announced its 2025 technical standards roadmap in January, specializing in specs addressing market modifications in privateness rules, information dealing with, and streaming media. The group plans to ship 31 new specs or updates all through 2025, in comparison with 23 releases in 2024.

Subscribe PPC Land publication ✉️ for comparable tales like this one


Subscribe

Timeline

Subscribe PPC Land publication ✉️ for comparable tales like this one


Subscribe

Abstract

Who: Anthony Katsur, CEO of IAB Tech Lab, a corporation that develops technical requirements for digital promoting with enter from over 800 member corporations globally. Katsur has labored in digital media for greater than 25 years.

What: Katsur warned that the promoting {industry}’s rush towards agentic AI represents peak hype whereas basic points together with provide chain transparency, privateness compliance, and measurement challenges stay unresolved. He emphasised protocols can not remedy issues attributable to misaligned enterprise incentives and unhealthy actors. The CEO expects agentic AI will deliver effectivity enhancements via pure language interfaces and automation however cautioned towards viewing it as an answer to structural {industry} issues.

When: The dialogue occurred on December 29, 2025, throughout what Katsur described as probably the most hectic yr in his 25-year profession in digital media. The dialog addressed developments all through 2025 and projected impacts extending via 2026 and past, together with the anticipated two-to-three-year innovation interval for agentic AI implementations.

The place: The context spans the worldwide digital promoting ecosystem, affecting publishers, advertisers, companies, and advert tech platforms. Particular challenges embrace fragmented privateness rules throughout 14 enforceable U.S. state legal guidelines in the beginning of 2025 with six extra anticipated all year long, European Union GDPR necessities, and evolving requirements affecting programmatic promoting infrastructure worldwide.

Why: The promoting {industry} faces a crucial resolution level between chasing agentic AI innovation and addressing foundational issues which have plagued programmatic promoting for years. Transparency points lead to solely 36% of post-transaction programmatic budgets reaching legitimate, viewable, measurable impressions in line with 2023 analysis. Katsur emphasised the necessity for privateness guard rails, measurement framework agreements, and provide chain transparency earlier than agentic techniques scale, warning that protocols alone can not remove dysfunction attributable to misaligned incentives between provide chain individuals. His place issues as a result of IAB Tech Lab develops technical requirements that decide how digital promoting infrastructure operates globally.


Source link