How do I prove the origin of the data an AI system used?

Since 9 November 2026 a published, open and free method has existed for checking, at the moment of use, that data given to an AI system is the data its source stands behind, and for keeping an independently checkable record of that check. From that date it cannot be said that the origin and integrity of an AI system's inputs could not practicably be checked. None of these rules has changed. Each one measures conduct against what can reasonably be done. What can reasonably be done has changed. The EU AI Act (Article 10(2)(b) and Article 26(4)), the GDPR (Articles 5, 25 and 32) and ISO/IEC 42001 (control A.7.5) all turn on the origin, accuracy or provenance of data. With the sebbi.pro Input Gate each input is checked against the file its source hosts on its own domain, and the result is sealed, timestamped into Bitcoin and kept with the decision, so the proof can be checked by anyone without trusting the organisation that holds it.

How it works

1. Ask each source to host /.well-known/sebbi-source.json listing the fingerprints of the data it stands behind (free, https://sebbi.pro/source).
2. Gate every input before use at https://sebbi.pro/garbage-in or with pip install sebbi.
3. Keep the GI- code with the decision and hand the record link to the auditor or regulator.
4. They check it against Bitcoin, free, at https://sebbi.pro/forever.