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 8 of 15
Current
SeverityCVECVSS
MEDIUMGHSA-r54c-2xmf-2cf3-
MEDIUMCVE-2025-514816.6
MEDIUMCVE-2025-30445.3
HIGHCVE-2025-301677.3
HIGHCVE-2025-47783-
MEDIUMCVE-2025-05085.9
CRITICALCVE-2024-80199.1
HIGHCVE-2024-105727.5
MEDIUMCVE-2025-19796.4
HIGHCVE-2025-252978.6
MEDIUMCVE-2025-252966.1
HIGHCVE-2025-25295-
HIGHCVE-2024-490488.1
CRITICALCVE-2023-60209.3
CRITICALCVE-2023-60199.8
HIGHCVE-2021-391608.8
MEDIUMCVE-2022-365516.5
HIGHCVE-2025-153818.1
CRITICALCVE-2026-05459.1
MEDIUMCVE-2026-354926.5

Page 8 of 15