In the first of three AI-specific predictions pieces, we asked an expert panel to gaze into the future of agentic AI, and how it will shape programmatic buying in 2026….

As artificial intelligence evolves from responsive tool to autonomous agent, programmatic advertising faces its most significant transformation yet. The emergence of agentic AI (systems capable of independent decision-making, goal-oriented behaviour, and complex task execution) promises to fundamentally reshape how digital advertising is bought, sold, and optimised.

But what does this shift truly mean for an industry already grappling with the implications of generative AI, escalating privacy regulations, and fragmenting media environments? 

To cut through the hype and identify the genuine opportunities and challenges ahead, we asked a panel of leading industry experts to share their predictions for AI and programmatic in the agentic age.

What the experts predict…

"Success will depend on collaboration."

This year saw the rise of AI agents, each impressive yet mostly confined within single platforms. While valuable for simplifying complex interfaces, their impact so far has been more evolutionary than revolutionary. The next major step will come as agent‑to‑agent communication, enabled by open standards like AdCP (Ad Context Protocol), becomes embedded throughout the programmatic supply chain.

As programmatic continues to evolve, the shift towards agentic systems represents its most exciting transformation yet. Autonomous buyer and publisher agents, supported by intelligent intermediaries, can negotiate, optimise, and act on behalf of their human counterparts. This progression from automation to true autonomy will free people to focus on creativity, strategy, and delivering meaningful outcomes for customers.

Success will depend on collaboration and a willingness to test, but the direction of travel is clear. The industry has never been more aligned or more ready to embrace the agentic age in ad tech.

Emma Newman, CRO EMEA, Pubmatic

"The industry will once again be forced to balance innovation with accountability."

2026 will push the media industry squarely into the "Peak of Inflated Expectations" for AI. Agentic applications will surge, especially in planning, creative automation, and measurement rather than media buying. There will be widespread testing and experimentation, but building dependable AI workflows will prove harder than promised. Failures will outnumber wins...

To unlock real scale, the industry will need far greater collaboration across the emerging standards taking shape, along with deeper involvement from the walled gardens to ensure interoperability and consistency.

And the elephant in the room will grow louder: how does sustainability keep pace? The industry will once again be forced to balance innovation with accountability, as it did with programmatic a few years ago.

Eric Shih, COO, Cedara

"Will the ends justify the well-intended efforts?"

There’s every chance this analysis of the various protocols and methodologies doing the rounds will be out of date by the time Jools Holland is scatting his way through the Hootenanny. Someone smarter than me has probably done the calculations, but agentic seems to have at least outlasted blockchain as the new frontier for trust and verification in the programmatic economy.

Whether it slows the collapse of the open web as a viable advertising channel at scale – itself a poster child for deeply flawed economics and inflated expectations – remains to be seen. If it can produce an ecosystem made up of environments that are used at a scale where (fewer, better) ads are at least tolerated enough to drive publisher revenue and tangible advertiser results, the ends will justify the well-intended efforts. If not, the money will continue to siphon out to Big Tech through little more than inertia.

Chris Andrews, Head of Marketing Technology, Miroma Founders Network

"The platforms that succeed will treat agents as transparent collaborators."

Agentic advertising is moving from experimentation to real application. The goal is the same as any other performance strategy: giving marketers predictable outcomes through clean supply.

By the end of 2026, the real breakthrough in AI and programmatic will not be creative automation. It will be the restructuring of buying decisions. Agents will become the default logic layer for planning, activation, and optimisation, evaluating supply paths, attention signals, and identity confidence in real time.

This reframes the auction from a race to the bottom into a negotiation around decision quality.

For SSPs, it means cleaner auctions, less noise, and supply paths that buyers can verify and act on. As agents scale, the ecosystem will need shared protocols for approvals and pricing logic.

The platforms that succeed will treat agents not as automation, but as transparent collaborators that optimise for outcomes without obscuring how decisions are made.

Amir Sharer, CEO, BRAVE

"2026 will move firmly toward revenue and real, demonstrable value."

2026 will be the year we start to see which AI frameworks and protocols truly take hold. A handful will emerge and fade, but by year end the industry will have a clearer view of which can thrive. The real question is which win and who ultimately owns and drives them. In many ways, owning the protocol is the tech equivalent of owning the narrative. 

Winning use cases will also become evident. 2025 has been dominated by excitement, hype, and proofs of concept, 2026 will move firmly toward revenue and real, demonstrable value. AI won’t be used as a blanket term; focus will shift to the specific value add. 

At the same time, it will be increasingly obvious that the programmatic stack can’t absorb another layer of cost. With SPO being increasingly in focus, AI and agentic solutions need to augment existing players or replace them, not add more middlemen.

Simon Kvist, CEO, Adnami

"The leaders will treat AI as a cross-team orchestration platform."

AI is already a powerful accelerator for creative and media teams; however, its rapid adoption is also widening the gap between them. In 2026, I predict a clear divide will emerge: organisations that use AI not only to speed up production, but to connect creative and media workflows, and those that don’t. 

The leaders will be the ones who treat AI as an orchestration platform, coordinating insights, content, and execution across teams. The laggards will be those who apply AI in silos, accelerating each function independently while missing the opportunity to bring them closer together.

Oz Etzioni, CEO & Co-Founder, Clinch

"Behavioural, contextual, and emotional data will merge into one predictive layer."

As we head into 2026, programmatic and performance marketing are converging into a new phase powered by predictive intelligence and cross-channel scalability. Growth will increasingly come from LATAM, SEA, and APAC regions, connecting global publishers with high-quality local demand. At the same time, the industry is shifting from fragmented media buying to unified, AI-driven decisioning across CTV, mobile, and web, where brands expect real incrementality, transparency, and measurable outcomes.

Predictive programmatic will move beyond segmentation to anticipating user intent, dynamically adjusting creatives, placements, and bids based on real-time signals. Behavioural, contextual, and emotional data will merge into one predictive layer, creating hyper-relevant experiences at scale. While AI accelerates optimisation cycles and opens new distribution opportunities, human oversight will ensure brand strategy and ethics remain intact.

Yaron Tomchin, CEO, Mobupps

"Agentic AI services will evolve from concept to early deployment."

As readily available LLMs lower traditional barriers to access, we’ll certainly see a surge of 'GenAI-first' platforms in 2026. That said, some incumbents will struggle to integrate AI meaningfully into their ecosystems - despite what their pitch decks might claim.

Agentic AI services will evolve from concept to early deployment, with some buying and selling agents likely to emerge through protocols like AdCP and UCP, but questions will remain about scalability until the technology matures and becomes accessible to smaller organisations. 

Meanwhile, the rise of GenAI, combined with a lack of commercial models for training LLMs, will reduce the supply of high-quality, human-written content, which publishers assume their users prefer and advertisers assume performs better. This will inevitably drive up CPMs, pushing more spend into walled gardens.

Matthew Beck, VP Strategy & Partnerships, Nano Interactive

"Businesses need to see AI operations as a key growth lever."

Are we entering the AI Agentic age? Absolutely. But this is just the beginning. The technology will undoubtedly become the standard way for most organisations to operate and there will be some at the cutting edge who have already embedded agents across core processes throughout their business, but they remain in the minority. The majority are still using AI in pockets to solve specific pain points, and these are not Agentic solutions in all cases. This siloed approach is the result of legacy organisational structures, deeply entrenched ways of working and deep-set behaviours. 

In order to accelerate the adoption of Agentic AI, businesses need to see operations less as a compliance function and more as a key growth lever with a huge scope for creativity. To overhaul a departmental operational model is a huge undertaking - it requires significant investment of time, focus and, of course, investment, but the benefits outweigh the initial pain points and will produce a net gain extremely quickly. Organisations that commit to a roadmap will reap the benefits whether that is licensing third party tech, building their own or a hybrid. By doing nothing you'll only fall further behind the competition.

Ben Foster, Chief Operating Officer, The Kite Factory