Indian patent application 202541119203 · published
Satyak signs photos and video on the phone itself, with hardware-bound cryptographic provenance that an AI generator can't forge. Real footage carries its proof. Fakes simply don't.
WHY PROVENANCE
Every new model makes yesterday's deepfake detector less reliable. Guessing after the fact never ends.
EXIF tags and watermarks can be copied, stripped or forged. Trust has to be anchored in hardware the forger doesn't have.
Authentic footage still captures bystanders. Satyak proves origin while protecting the faces that didn't sign up to be seen.
HOW IT WORKS
A photo or video is taken on the phone's physical camera sensor.
A hardware-evident signature binds the media to the device, something AI without physical access can't produce.
Configurable identity-based encryption hides chosen faces, independently of the authentication layer.
An edit chain keeps proof intact through permitted edits, so anyone downstream can verify.
Under 85 ms overhead in authentication-only mode on flagship smartphones.
FIVE PRIVACY MODES
WHERE IT MATTERS
Publish field footage with proof it came from a reporter's camera, not a generator.
Give photo and video evidence a verifiable chain from capture to courtroom.
Accept damage photos with confidence they were shot on site, today.
Stop injected and synthetic selfies at onboarding with capture-time proof.
Verify citizen and field reports while keeping bystanders' faces protected.
PROTECTED INNOVATION
Satyak's capture authentication is the subject of a published Indian patent application. Privacy controls are designed into the capture pipeline itself, in step with India's DPDP framework, rather than added after the fact.
Mobile device capture-time authentication system for deepfake-resistant media using hardware-bound cryptographic provenance and configurable selective identity-based encryption
We're onboarding pilot partners in media, legal, insurance and identity. Tell us about your workflow.