Profila Sign up

Profila Research Labs

Advertising that proves
a match without
revealing the person.

Research into serving relevant offers under cryptographic guarantees — the counterparty learns that a match exists, never who.

Zero-Knowledge Advertising Published

2026

Zero Knowledge Advertising: a new era of privacy-preserving AdTech solutions

Callejo, Cuevas & Cuevas (UC3M) with Kotila, Bragg & Van Roey (Profila).

A simple, efficient privacy-preserving advertising solution that combines the strengths of Brave, FLoC and Personal Data Platforms. Ads are personalised and delivered without sharing any user data with third parties (zero-knowledge), with an optional consent mode for fine-grained data sharing. Simulations show it runs within the ad ecosystem’s timing budget (tens of milliseconds) while using under 2.5% CPU and under 0.1% memory — viable on battery-powered handheld devices.

Profila Research Labs

01 · How it works

Zero-knowledge by default. Consent when you choose.

In the default mode, ads are personalised and delivered without sharing any of your data with third parties. A second, opt-in mode lets you hand a specific advertiser a specific fact — your size to a clothing brand, your location to a restaurant, your taste in music to a concert promoter — and be rewarded for it. You choose what, and with whom. Nothing else moves.

Profila Research Labs

02 · The mechanism

The match happens on your device.

A zero-knowledge ad server sends your device an encrypted pool of candidate ads. Your device picks the relevant one locally and renders it — the server never sees your data, and never learns which ad matched. A peer-to-peer network routes every message so your IP address is never exposed either. Under the GDPR an IP address is personal data; here it never leaves the network.

Profila Research Labs

03 · Why it's more relevant

Declared, not inferred.

Today’s AdTech builds a profile by guessing from your browsing history — and inference has been shown to be inaccurate. This solution is active: you declare what you’re interested in, so the ads you see are strongly aligned with what you actually said, not with what an algorithm assumed about you.

Profila Research Labs

04 · The guarantee

Every step leaves a proof.

Granting consent, revoking it, delivering an ad, settling the bill — each action generates a cryptographic, non-repudiable proof, stored in an auditable repository (a blockchain with smart contracts). If an advertiser keeps using your data after you revoke consent, there is undeniable evidence of it. The privacy guarantee is auditable, not just promised — by you, and by third parties.

Profila Research Labs

05 · Is it practical?

You’re paid — and it runs on your phone.

Users are compensated for interacting with ads and for any data they choose to share. And the design is deliberately light: simulations deliver ads in tens of milliseconds — inside programmatic advertising’s few-hundred-millisecond budget — while consuming under 2.5% CPU and under 0.1% memory. It works on battery-powered handheld devices, not just servers.

Profila Research Labs

How it stacks up

The only approach that is all six at once.

Zero-knowledge Consent Active Rewards Auditable
Brave
Google FLoC
PDPs
This solution

After Table I. The comparison frames alternatives that can coexist, not a ranking.

Profila Research Labs

In one line

What it asks

Can advertising be personalised and settled without any party — advertiser, exchange, or anyone else — learning who the person is? The paper’s answer is yes, and it proves it cryptographically. The privacy guarantee becomes a proof, not a promise.