GumGum launched Mindset Agent on July 15, an AI tool that turns a campaign brief directly into a custom targeting segment inside its redesigned Mindset Portal, with The Trade Desk as the first activation partner. The agent runs on GumGum’s Mindset Graph, which the company says processes billions of contextual, creative, attention, environmental and historical signals through a vector-based model to match content and audience intent without relying on cookies or other personal identifiers. HEINEKEN, e.l.f. Cosmetics and the BBC are named as early users.
The launch matters beyond one vendor’s product sheet. As third-party cookies keep eroding and identity-based targeting gets harder to justify to privacy-conscious marketers, contextual intelligence platforms are racing to prove they can do more than block-list keywords. Turning a plain-language brief into an activatable segment, without a planner manually building it, is GumGum betting that contextual data can carry the same automation the ad industry has spent a decade building around cookies and device IDs.
The original insight here is less about GumGum specifically and more about where the “AI agent” label is landing in ad tech: not just on bidding and creative generation, but on the planning step that used to require a human strategist reading a brief and translating it into targeting logic. “AI should not just automate old targeting models,” said GumGum CEO Marcus Startzel. “It should help brands make better decisions about where, when, and how to show up.” If that holds, the next competitive line in cookieless advertising will not be who has the most contextual data, it will be who can turn a brief into a live segment fastest.
Related: CTV’s Premium Price Tag Comes With a Trust Problem and Attention Measurement Gets a Pre-Campaign Metric track the same shift toward pre-campaign AI tooling and inventory scrutiny.
Source: GumGum
