Continuous
Measurement at a 15-minute cadence. Never a stale snapshot.
Anti-gaming
Verdict = minimum over a rolling 24 hours. No one can game the signal.
Integrity proof
SHA-256 hash-chain + anchoring + RFC 3161 timestamping via a qualified eIDAS TSA. Verifiable by a third party.
Neutral
Paid by the buyer, never by the measured party. No conflict of interest.
Illustrative example — timestamped verdict
AI system "IA-203" · 8 model versions hashed · 3 training datasets sealed · complete timestamped journal.
Is my AI system high-risk?
The AI Act defines categories of high-risk systems (health, education, critical infrastructure, recruitment, etc.) subject to reinforced obligations. Check Annex III of the regulation; 0DATA documents compliance, it does not qualify it.
What are a provider's AI obligations?
Technical documentation, risk management, data traceability and logging, especially for high-risk systems. Sealed, timestamped proof documents these obligations in a verifiable way.
How do I document a high-risk AI system?
By sealing each model version and each training dataset (SHA-256 hash), eIDAS-timestamped and anchored on Hedera. You prove the provenance and integrity of your system at any point in time.
What does non-compliance with the AI Act risk?
Fines up to 7% of global turnover depending on severity. Continuous sealed documentation reduces risk by demonstrating the required traceability.
How do I prove the traceability of my training data?
By anchoring the hash of each training dataset and each model version. The eIDAS-timestamped anteriority fixes the date of your data, verifiable by a third party.
Does the AI Act apply to my use case?
It depends on your role (provider, deployer) and your system's risk level. 0DATA documents technical compliance, but legal qualification belongs to your counsel.