LIKENESS & VOICE · PRODUCT VALIDATION
Your face. Your voice.
Your say.
IN VALIDATIONA proposed watch for name- and handle-linked impersonation, reviewed against consented reference media. No customer monitoring or biometric search today.
No charge · no launch-date promise
THE PRODUCT BOUNDARY
Your identity stays yours.
Discovery starts from identifiers you intentionally register.
Consented photos, video, and voice help review what the search finds.
No face or voice identifier is computed to discover strangers.
- No customers watched today
- No AI-generation claims until the detector is validated
- Name-free impersonations are outside scope
BEST FIT
Who it's for
- A founder or creator whose name fronts the brand, and who wants suspected impersonation organized by the name and handles it trades on, never by biometrics.
- A public-facing expert or executive whose endorsement carries weight. This tier is built to organize source-linked evidence around a suspected fake endorsement for human review.
- A team that has already fielded an “is this really you?” message and wants a future watch to preserve the source and rationale when a name-linked candidate appears.
VALIDATION WAITLIST
Help shape what comes next.
PRODUCT STATUSWhat works, what does not, and privacy requirements
THE RELEASE BOUNDARY
The full-surface target claim, shown as a target rather than present coverage:
We monitor YouTube, podcast directories, app-store listings, and the Meta Ad Library (EU/UK commercial ads) for content using your name and flag AI-generated audio/video among the matches. We do not perform biometric face or voice search.
The by-design miss, stated rather than papered over: Impersonations that never use your name are not visible to this watch.
What works in validation
The validation components accept an exact name or handle, preserve submitted source evidence, and can compare supplied candidate media with reference photos, video, or voice supplied with consent. Reference media remains ordinary stored media; no biometric identifier is computed.
We do not perform biometric face or voice search. That sentence is true today and stays true by standing rule: no biometric processing of anyone's face or voice enters this product without a deliberate privacy-policy change and a new review of the customer promise.
What is not covered today
No customer monitoring is running and this product is not available for purchase. An app-store search has completed a real validation query; YouTube and podcast access are not connected. AI-generation detection is also not connected, so MarkSentry makes no AI-generation claim. Impersonations that omit the watched name or handle remain outside this approach.
What we require before release
At least two named official surfaces must return source-linked evidence, the selected detection service must operate within the stated privacy boundary, and a real finding or documented no-hit must retain evidence that can be independently opened. Enrollment, billing, retention, consent, and the zero-biometrics rule must all be verified before sale.
What the reference vault stores, and how long it is kept, is set out in the privacy policy.
Why we're building it this way
Official source → evidence trail → your call
MarkSentry begins with sources it can name and evidence it can preserve. Any unreleased commerce or open-web watch remains unavailable until its access, matching, and evidence trail are validated.
Discovery is limited to your exact name and handles on named official surfaces. Candidate media may be reviewed for reference-media reuse; AI-generation classification remains a validation target. One rule does not change: We do not perform biometric face or voice search.
Built to organize name- and handle-linked impersonation evidence for review, using consented reference media and never biometric identification.