How IdGenum Works

Registration starts with a short biometric enrollment session. You capture a brief video of your face using your device camera. That video never leaves your device in raw form — it is transmitted directly to an AWS Nitro Enclave, a hardware-isolated processing environment with no network access and encrypted memory.

Inside the enclave, your facial geometry is analyzed and converted into a BioHash — a non-reversible cryptographic hash generated using Index-of-Maximum (IoM) random projection. The raw biometric data is destroyed immediately after hash generation. It is never stored anywhere.

Your BioHash is then encrypted with a personal KMS key and stored securely. At the same time, a blockchain-anchored Decentralized Identifier (DID) is created for you — this is your identity on the public ledger that you own, not us. A W3C Verifiable Credential is issued as your official certificate of registration.

Once registered, automated detection begins scanning the web for unauthorized uses of your likeness. Scanning frequency depends on your plan — weekly for Pro subscribers, twice weekly for Creator Pro.

Detection uses a multi-vendor approach for accuracy. Multiple independent deepfake detection services analyze content simultaneously. Both vendors must agree before a match is confirmed — this dual-confirmation model minimizes false positives and protects you from incorrect takedowns.

When a potential match is found, evidence is automatically collected: the source URL, a timestamped screenshot, confidence scores from each vendor, and a biohash match score. All evidence is written to a tamper-evident, cryptographically hash-chained record with write-once storage — creating a legally admissible chain of proof.

When unauthorized content is confirmed, IdGenum generates evidence-backed takedown notices fully compliant with the NO FAKES Act. Each notice includes six legally required fields: identification of the unauthorized replica, identification of you as the right holder, evidence of your prior registration, a good-faith statement, an accuracy statement under penalty of perjury, and designated agent contact information.

Notices are only generated when combined detection confidence reaches 95% or above — both vendors must agree. This high threshold exists because the NO FAKES Act imposes a $25,000 penalty for false takedown notices. Your reputation and legal standing are protected by design.

Platforms that receive a valid notice must respond within 10 business days. If they do not comply, they lose safe harbor protection and face penalties up to $750,000 per work. You maintain full visibility into the status of every takedown through your violation dashboard.