• by Laurie Sullivan , Yesterday

DoubleVerify has released a "cognitive engine" that powers artificial intelligence-based (AI) agents across its platforms to carry out agentic tasks.

The cognitive engine bridges the gap between ad-verification data and campaign execution. It allows agents to carry out agentic tasks including campaign analysis, optimization, and activation that support clients’ media buying decisions and workflows. 

The engine, “DV Neura," powers specific features to enable communication between third-party AI agents and DoubleVerify.

Butler/Till, an independent marketing and media agency, became an early adopter of the agentic technology, which allows agents to conduct a variety of tasks, including buying media.

DoubleVerify CEO Mark Zagorski described the cognitive engine to MediaPost as a “brain” that ties together a variety of AI-powered tools across DoubleVerify's platform.

The company says DV Neura is organized around four structural principles or areas of focus that include “media intelligence,” “adaptive performance,” “open connectivity” and “agentic execution.”

Adaptive performance covers bidding optimization and campaign measurement, whereas open connectivity defines how external systems and AI tools access DV's data.

Agentic execution is the layer that connects insight to action. Zagorski says the dashboard is no longer needed.

In response to the question of whether agentic AI changes the company’s business model, Zagorski said it changes the way advertisers and their agency partners interact with DoubleVerify.

“Rather than go to a dashboard to see how their campaign is doing, the technology gives advertisers the ability to query the chatbot to get updates on brand suitability, for example, through natural language queries,” Zagorski said. “It moves the relationship from a push to a pull model,” he said.

In a classic model, brands and agencies must pursue the answers. This evolution into agentic allows a platform to bring the answers to brands and agencies before they begin to search.

AI "slop" is still one of the biggest issues that advertisers must contend with. Media Intelligence supports fraud detection and content classification, but it also addresses what Zagorski calls advanced AI to filter out slop, low-quality, mass-produced generative AI content, as well as stops the delivery of fraudulent impressions.

“AI slop has exploded,” he said. “The amount of bad content is clogging the advertising channels. It’s massively produced.”

DV Neura Insight Agent is available through the Model Content Protocol (MCP) and Anthropic Claude integration. The DV Neura Activation Agent, which autonomously executes approved campaign changes within an advertiser’s guardrails, is scheduled for release in the third quarter of 2026. 

The company plans to make the protocol available through all AI platforms in the future, with additional MCP integrations becoming available with Google Gemini and Microsoft Copilot in the near future.