Amir Sharer, founder & CEO of BRAVE, chats to four industry experts to discuss their thoughts on the new programmatic ecosystem in 2026...

Over the last five years, programmatic predictions have focused on software, including new algorithms, ID solutions, and creative tools.

But looking at 2026, the conversation has shifted. As agentic AI becomes the dominant media buyer, the current infrastructure of the open web is buckling under the weight of billions of micro-decisions and a rising tide of AI-generated content.

We are entering The Great Reconstruction.

The market has already begun consolidating. According to ANA’s Q2 2025 Programmatic Transparency Benchmark, curated marketplaces and PMP transactions now represent the majority of programmatic spend. The shift toward controlled, higher-signal supply is no longer theoretical.

The winners in 2026 will not be the platforms with the most access. They will be the ones with the strongest filters, the fastest pipes, and the cleanest signal. I sat down with four industry leaders to map this new terrain. Here are the architectural shifts defining 2026.

Agentic Logic Rewrites the Rules of Programmatic Scale

The era of 'spray and pray' is officially over. Agentic AI doesn't just buy media; it negotiates outcomes. This requires a completely new protocol to function.

Brian O'Kelley, CEO of Scope3, puts the magnitude of this shift into perspective: "AdCP (Advertising Content Protocol) and agentic advertising are opening up hundreds of billions worth of new opportunities to marketers and content creators. AdCP is the most significant new advancement... It will fundamentally change the way we think about programmatic advertising."

Lisa Abousaleh, SVP partnerships at ID5, notes that we have crossed a threshold where AI systems now run the workflow: "In 2026, AI won’t just support programmatic workflows; it will increasingly run them. Instead of humans configuring rules... advertisers will rely on AI systems that continuously learn, predict, and act in real time.

Lisa adds that these protocols provide a "structured, machine-readable way for buyers to communicate objectives... allowing AI agents to translate strategy directly into execution."

If Agentic AI is the engine, latency is the fuel. These agents negotiate outcomes in milliseconds, and speed becomes a competitive variable. Infrastructure built on heavy abstraction layers will struggle to compete with architectures designed for real-time execution.

Just as viewability and brand safety became contractual requirements, infrastructure performance is moving in the same direction. In 2026, latency expectations will increasingly become part of the buying conversation. For high-frequency mobile environments, response times below 10ms are increasingly becoming the benchmark for competitiveness.

In an environment where auctions close in milliseconds, slow response times don’t just reduce efficiency. They reduce competitiveness.

Curation Moves Upstream (The End of "Optimisation")

For a decade, the workflow was "Launch, then Optimise." That cycle is now too slow.

Terrence Moduthagam, VP & head of Media.net Select, argues that intelligence is moving to the planning phase: "The teams that see the best results are being more intentional upfront... instead of relying on optimisation to fix things later."

He notes that API-first activation is the critical unlock here, allowing sell-side curation to finally become "scalable, automated, and measurable across the open internet."

Lisa Abousaleh calls these "goal-based, machine-led workflows": "Humans define intent, and AI handles activation... To prepare, companies need to shift from channel-centric and signal-centric thinking to outcome-focused, signal-agnostic systems."

We are seeing the erosion of the generic SSP model and the rise of Curation-as-a-Service. Buyers are no longer interested in accessing all of the supply: they want pre-packaged, format-specific bundles that align with user consumption before the first bid is placed.

The Identity Paradox: Targeting the Machine to Reach the Human

Targeting in 2026 will increasingly mean targeting decision systems, not just demographics. 

Lisa Abousaleh identifies a new agentic intermediary that complicates targeting: "Advertisers won’t just be trying to understand people; they’ll be trying to understand people plus the agents representing them... Who is the audience: the individual, the household, or the agent optimising on their behalf?"

Lisa calls for "adaptive technology," and we agree. In an era where a buying agent might make the purchase decision on behalf of a human, "signal fidelity" becomes the primary pricing lever. Infrastructure must be able to ingest and validate these fragmented signals (deterministic IDs, probabilistic models, and agent protocols) without slowing down the pipe. If your architecture relies on perfect visibility, your team will be blind in 2026.

To solve this, Tamir Shub, VP business development at Intent IQ, predicts a pivot toward "self-optimising ecosystems": "2026 marks a pivot... integrating agent logic into first-party identity graphs that will lead to more effective identity matching and better monetisation."

The Trust Shift: Selectivity as a Strategy

As automation scales, so does opacity.

Terrence Moduthagam warns that the biggest challenge in 2026 is complexity: "It’s getting harder to understand what’s actually driving results, especially as traditional targeting methods continue to change." 

Amir Sharer, founder & CEO, BRAVE: "This complexity drives a flight to safety. At BRAVE, we believe that the technologies driving innovation this year are not the ones making programmatic 'louder': they are the ones making it more selective, explainable, and trustworthy. We are seeing a flight to safety where features like OM SDK adoption and pre-bid measurement are not just 'nice-to-haves'. They are the entry fee."

In 2026, uncertainty becomes the most expensive line item. Buyers will pay a premium to avoid it. The winners will not be the loudest platforms, but the ones that make the most trustworthy decisions at scale.

The Rise of the "Human CPM"

As the open web floods with AI-generated content, human attention is becoming a scarce luxury good.

Terrence Moduthagam identifies this as the moment programmatic finally aligns with "how users actually consume content," specifically citing formats like vertical video moving from experimental to "foundational."

At BRAVE, we predict a bifurcation of the market:

• Tier 1: verified human environments (gaming, authenticated apps, vertical video).
• Tier 2: unverified open web environments

Buyers will stop optimising for the lowest CPM and start optimising for the Human CPM: media valued not by volume, but by verified human presence.

The Human CPM will be defined by measurable signals: authenticated app environments, verified session depth, OM SDK instrumentation, and pre-bid fraud filtration. If an impression cannot demonstrate a real human context, it will be discounted or excluded altogether.

The Blueprint for 2026

The ad tech industry is not shrinking. It is fortifying.

As Brian O'Kelley noted, this shift is "opening up hundreds of billions worth of new opportunities." But capturing them requires an infrastructure paradigm shift.

Do not invest in bigger pipes. Focus on smarter filters. Do not aim for more scale. Focus on better, more accurate signal.

Welcome to the Reconstruction.

Brave

BRAVE is a radically human Supply-Side Platform (SSP) engineered for precision, performance, and transparency across mobile, CTV, and web. By combining AI-driven curation, predictive supply path optimisation, and signal-rich monetisation, BRAVE h...

More about Brave »

Powered by PressBox