Product & technology

Introducing AI Insights: Data clean room results in plain language

Written by

David Sturzenegger

Published on

October 1, 2026

Two translucent glass bars, one teal and one purple, cross in an X with a glowing orange block where they overlap, against a dark background.

Decentriq's new AI Insights turns the statistics in a Decentriq data clean room into a plain-language brief that describes who is in the overlap audience and which segments to prioritize or avoid. Marketers can then turn the suggested segments into a rule-based audience in a few clicks. AI Insights ships as part of platform release v5.12.

Request a live demo

Want to see what else data clean rooms can do? Have a specific use case in mind? Let us show you.

Request demo

A woman with dark hair sitting on a city rooftop reading something on a tablet

Faster answers from clean room insights

Insights in a Decentriq data clean room show how an advertiser's seed audience overlaps with a partner’s audience, such as a publisher or retailer, broken down by age, gender, interests, and affinity. The level of detail is what makes them valuable, but it is also what can make them slower to read, interpret, and ultimately act on.

A marketer has to scan several charts and decide which segments are worth acting on. Many teams stop at the overlap number. AI Insights closes that gap by doing the first read for them.

What's new: an AI Summary on the Insights tab

Every data clean room with AI Insights enabled now has an AI Summary panel at the top of the Insights view, with the supporting charts alongside. It reads the aggregated results and writes up:

The top-line read: a short description of who the overlap audience is and how it differs from the publisher's wider base.
Where the lift is: the segments that over-index most strongly, for example a given age band combined with an interest category, with their affinity.
What to do and what to avoid: a short list of segments worth prioritizing, and a separate list of segments that under-index and are better left out.

The summary is anchored on the data in the clean room. Where the signal is weak, it keeps the recommendations short.

Example AI insights: The best travellers aren't in the Travel section

Turning a summary into an audience in a few clicks

Each attribute combination named in the summary, such as an age band paired with an interest category, can be selected directly. Once a selection is made, Decentriq opens the rule-based audience builder with it already set up: combinations with an affinity above 1 are included, and those below 1 are excluded.

Criteria do not need to be re-entered by hand, so a marketer can go from reading the summary to a first audience definition in one flow. For teams that want more reach, the Insights view also points to AI Lookalike audiences as the next step.

Built on aggregated results, with customers in control

AI Insights works only from the aggregated results a data clean room already produces: the overlap statistics and insights that participants see today. These outputs are aggregated and contain no personal data. The underlying first-party data stays inside the confidential computing environment and is never sent to the AI.

Sharing aggregated results is what switches AI Insights on, and customers can opt out. Every summary is clearly labeled as AI-generated, so teams know to treat it as a starting point and check it against the charts next to it.

Get started

AI Insights is available now in Decentriq platform release v5.12 for media and advertising data clean rooms. The AI Summary appears on the Insights tab of the clean room. Customers who want to enable AI Insights or opt out can contact their Decentriq customer success manager or get in touch with the team. A demo is also available: request a demo.