The rollout of ChatGPT Ads, in the U.S. in February 2026 and in Europe in late August, marks a historic turning point.

For the ad industry and for VCs, the moment echoes the birth of Google AdWords and the rise of programmatic. Traditional search monetized intent captured in a single keyword. Generative AI search engines handle an iterative conversation, a full semantic continuum.

For early-stage tech investors, this shift is more than a technical adjustment. It's the basis of a new investment thesis and an unprecedented wave of opportunity for the next generation of startups.

The market opportunity is huge. Generative AI is on track to become the fastest-growing ad channel in history, heading toward $100 billion by 2030, according to June 2026 figures from WPP Media. As chat interfaces become people's front door to the open web, where they browse and make decisions, ad formats have to work for humans and bots alike.

In the agentic era, advertising has a new target. The goal is no longer to capture human attention but to influence a machine's decision.

That shift locks in the merger of search, e-commerce and retail media inside the prompt.

The challenge is infrastructure. How will legacy players – DSPs, SSPs, ad exchanges, ad servers – adapt? How do you price a fraction of a conversation in real time? The impressions-and-clicks model is giving way to a semantic exchange that surfaces the single right product. Which ad formats, content and catalogs will get brands seen and, more importantly, understood by bots?

For VCs, the agentic era kicks off an exciting new investment cycle.

Look at audience building (Webedia), video programmatic (StickyADStv), mobile (Adikteev, Ogury), connected TV (Olyzon), attention measurement (Xpln) and brand visibility in agentic commerce (Botify). The pattern holds every time: Each technology shift weakens entrenched incumbents and hands value to nimble newcomers.

Up against the foundation model giants (OpenAI, Google AI Overviews, Claude, Perplexity), startups won't win by reinventing the underlying models. Their opening is in the middleware and application infrastructure layer. A whole suite of critical tools still needs to be built from scratch:

LLM optimization (LLMO) and generative engine optimization (GEO): Brands badly need tools to manage and audit how they show up in model answers.

Attribution and semantic influence measurement: How do you quantify the business impact of a brand mention buried in an unstructured conversational summary?

Semantic "ad exchanges" and agent-to-agent protocols: real-time matching of autonomous shopping agents' requests with merchants' catalogs.

Conversational brand safety and compliance: making sure in milliseconds that a brand recommendation isn't tainted by hallucinations or toxic content.

The economics of this new wave favor early-stage funds like Ventech. Ultra-lean, AI-native teams of two or three people now build highly scalable deep-tech software. That's a clean break from the services models and heavy cost structures of the past.

The edge will always go to founders who pair a deep understanding of the models with sharp industry expertise and a broad network of publishers and brands. Network effects will remain the barrier to entry. They were key to StickyADStv's success: Comcast bought it for its lead with European video publishers.

In a global industry where the U.S. still calls the shots, generative AI and the agentic era are ushering in one of the most promising investment cycles in ad tech history.

The next generation of ad tech unicorns won't monetize banners or search ads. They'll build the infrastructure for semantic commerce worldwide.

By Claire Houry, general partner at Ventech

Keep reading