Taboola CEO Adam Singolda says AI is remaking ad discovery at POSSIBLE

At POSSIBLE, Adam Singolda, CEO of Taboola, framed the company’s platform shift as less about tinkering with ads and more about rebuilding the “match” itself using AI. In other words: Taboola is not only distributing recommendations or running paid discovery. It is betting that AI will increasingly decide what gets shown, to whom, and why that experience is relevant enough to earn attention and performance.

That matters because recommendation advertising lives and dies on one thing: relevance at scale. Taboola’s core business model depends on connecting advertisers with audiences through content-based discovery, typically mediated by user intent signals and engagement patterns. When AI is brought in at the center of that system, it changes the practical mechanics of the platform. Instead of treating targeting as a relatively static set of rules, the system can adapt to ongoing behavior and content context, potentially improving both the advertiser outcomes and the publisher experience. The reason Singolda’s message is worth executives’ attention is simple. Once AI becomes the engine of delivery, you are no longer buying reach alone. You are buying the intelligence that converts that reach into results.

AI transformation in ad tech also arrives with a board-level subtext: governance, data strategy, and platform risk are moving upstream. AI-driven advertising platforms rely on large sets of behavioral, contextual, and performance data. That means questions that used to live in engineering now land on the desk of compliance, finance, and risk committees. What data is used? How is it processed and stored? How is quality controlled? And what happens if models drift, or if measurement changes because the system changes how users and content are selected?

Even if the details of Singolda’s comments at POSSIBLE are framed around product evolution, the broader market context is hard to ignore. Across digital advertising, the biggest pressure on companies is that every improvement in targeting and personalization has to be balanced against user trust, privacy expectations, and increasingly strict regulation. The AI layer can amplify performance gains, but it can also amplify compliance burdens if companies cannot clearly explain and control how decisions are made. For executives, that means the AI roadmap is inseparable from policy readiness. A recommendation platform that upgrades its matching with AI may also need stronger auditability and more disciplined experimentation, because outcomes can change quickly when the selection logic changes.

There is another second-order implication that matters for Taboola and peers: AI can compress differentiation. When multiple platforms adopt similar machine learning approaches, performance gains can become less about a single proprietary model and more about the surrounding system. That includes feedback loops, experimentation cadence, the quality of training signals, and how quickly a platform can learn from new formats and new advertiser objectives. In practice, executives should assume that AI will raise the baseline. The competitive advantage shifts from “who has AI” to “who operationalizes AI best,” including how reliably the system delivers outcomes and how efficiently it turns learning into product improvements.

That operational reality is where the incentives get interesting. Advertisers want predictable, measurable return. Publishers want relevance that keeps audiences engaged without degrading experience. Taboola’s platform positioning, as presented by its CEO at POSSIBLE, suggests the company believes AI can satisfy both sides by improving discovery relevance. But the governance question remains: if AI decides what is shown, the company needs to ensure that the incentives of different parties do not collide. For example, optimizing purely for short-term engagement could conflict with long-term user satisfaction, while optimizing for advertiser conversion might change how content is surfaced for publishers. AI might make these trade-offs more dynamic, but it does not remove them.

So what should decision-makers take away from Singolda’s framing? AI is transforming Taboola’s platform in a way that shifts value creation from manual optimization to continuous, model-driven matching. For boards and leadership teams, the strategic stakes are about making sure that the AI transformation is accompanied by disciplined measurement, robust compliance planning, and clear accountability for how platform decisions impact user experience. In a world where AI increasingly governs discovery, the companies that win are likely the ones that can scale learning without losing trust or control.