AI Component

Training Data

Training data is both the model's most valuable input and its most underprotected one. Three problem classes dominate. First, poisoning: an attacker who can influence a public dataset, a web crawl, or a fine-tuning corpus can plant backdoors or biases that survive into the deployed model — BadNets-style attacks on image classifiers, trigger-phrase attacks on LLMs, and reward-hacking on RLHF datasets. Second, memorization and leakage: models can regurgitate verbatim training data, exposing PII and copyrighted content; this has driven the active New York Times v. OpenAI litigation and is a recurring GDPR concern. Third, provenance: when training data origins are unclear, downstream users inherit legal and security risk they can't assess. EU AI Act Article 10 (Data Governance) and ISO 42001 Annex A treat training-data quality as a controlled asset. Defenses: data lineage tracking, deduplication, PII scrubbing before training, and adversarial training against known trigger families.

287
Total CVEs
15
Pages
Page 11 of 15
Current
SeverityCVECVSS
MEDIUMCVE-2026-536566.3
HIGHCVE-2026-54058-
HIGHCVE-2026-277758.8
MEDIUMCVE-2026-584355.4
HIGHCVE-2026-648328.8
MEDIUMCVE-2026-650106.6
MEDIUMCVE-2026-660076.5
HIGHCVE-2021-478168.8
CRITICALCVE-2026-687719.8
HIGHCVE-2026-712818.8
MEDIUMCVE-2026-122615.3
HIGHCVE-2026-189478.5
MEDIUMCVE-2026-189425.5
CRITICALCVE-2026-189489.9
HIGHCVE-2026-189417.7
UNKNOWNCVE-2026-20728-
UNKNOWNCVE-2021-33627-
UNKNOWNCVE-2022-24069-
MEDIUMCVE-2026-28707-
HIGHCVE-2026-751117.5

Page 11 of 15