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 9 of 15
Current
SeverityCVECVSS
MEDIUMCVE-2026-18396.5
HIGHGHSA-89gg-p5r5-q6r47.7
HIGHCVE-2026-68598.8
HIGHCVE-2026-41486-
HIGHCVE-2026-40171-
UNKNOWNCVE-2026-31249-
UNKNOWNCVE-2026-31250-
HIGHCVE-2026-312537.3
HIGHCVE-2026-26147.5
CRITICALCVE-2026-312149.8
UNKNOWNCVE-2026-31218-
UNKNOWNCVE-2026-31219-
HIGHCVE-2026-312228.8
HIGHGHSA-7j65-65cr-6644-
HIGHCVE-2026-26528.6
HIGHCVE-2026-87567.3
HIGHCVE-2026-436248.2
HIGHCVE-2026-54228.1
LOWCVE-2026-108013.6
LOWCVE-2026-108033.6

Page 9 of 15