Notifications
Clear all
Topic starter
10/07/2026 9:17 pm
TL;DR: A scan of 7.6 petabytes across 187 million public AI training files found 221,303 live, unique credentials in 6,003 datasets, including write-capable tokens and cloud keys that can affect software, infrastructure, and identity systems, according to TruffleHog. The security problem is not only leakage, but permanent propagation across derivative corpora, which turns one exposed secret into an enduring governance failure.
NHIMG editorial — based on content published by TruffleHog: Scanning 7.6 Petabytes of AI Training Data for Secrets
By the numbers:
- The scan covered 7.6 petabytes across 187 million files.
- One Infura key got copied into 1,131 public datasets and 10,162 file locations.
Questions worth separating out
Q: What breaks when live secrets are published inside AI training data?
A: The normal secrets lifecycle breaks because the credential stops existing in one place.
Q: Why do public training datasets create more risk than a normal source-code leak?
A: Because training datasets are intentionally replicated and redistributed.
Q: How do security teams know which exposed credentials in datasets matter most?
A: Start with the authority the credential grants, not the system where it was found.
Practitioner guidance
- Scan corpora before publication Run secret detection across any dataset before release, including archives, notebooks, JSONL, and derived artefacts.
- Classify credentials by authorised action Prioritise tokens and keys that can push code, rewrite workflows, administer organisations, or push images.
- Assume downstream replication until proven otherwise Create incident runbooks that treat every public dataset copy as an independent exposure point.
What's in the full report
TruffleHog's full analysis covers the operational detail this post intentionally leaves for the source:
- The verification method used to confirm which credentials were live at scan time.
- The dataset families and model corpora with the highest concentration of exposed keys.
- The breakdown of credential types by cloud, database, AI provider, and publishing access.
- The practical revocation approach used when one secret appeared across many derivative datasets.
👉 Read TruffleHog's analysis of secrets hidden in 7.6 petabytes of AI training data →
AI training data secrets: what IAM teams need to act on?
Explore further
10/07/2026 9:40 pm
Training data secret exposure is now a governed identity problem: This is not just data leakage, it is credential distribution at scale. Once a live secret enters a public training corpus, it becomes an NHI asset embedded in a replication pipeline that organisations do not control. The implication is that AI data governance and secrets governance can no longer be separated.
A few things that frame the scale:
- The average time to mitigate a leaked secret is 36 hours, highlighting the operational burden of manual remediation processes, according to the 2024 State of Secrets Management Survey.
- 54% of organisations are dissatisfied with their current secrets management solution because not all secrets are secured, and 43% cite lack of central management.
A question worth separating out:
Q: What should organisations do when they find a live key in training data?
A: Revoke it, then search for every derivative copy that might still contain the same secret. If the credential was scraped into public datasets, the risk is distributed and the response has to be distributed too. Treat the incident as a lifecycle failure, not a one-off leak, and confirm invalidation across all known copies.
👉 Read our full editorial: Training data secret exposure is now an NHI governance problem
