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 5 of 15
Current
SeverityCVECVSS
LOWCVE-2025-42873.3
MEDIUMCVE-2025-461525.3
MEDIUMCVE-2025-461535.3
MEDIUMCVE-2025-536216.9
HIGHCVE-2024-340727.8
HIGHCVE-2025-69217.5
CRITICALCVE-2025-626089.1
HIGHCVE-2022-07367.5
CRITICALCVE-2023-11779.8
HIGHCVE-2023-301727.5
CRITICALCVE-2023-27809.8
CRITICALCVE-2023-376510.0
HIGHCVE-2023-40337.8
HIGHCVE-2023-60157.5
CRITICALCVE-2023-60189.8
CRITICALCVE-2023-60149.8
HIGHCVE-2023-434727.5
HIGHCVE-2023-67098.8
HIGHCVE-2023-67538.8
HIGHCVE-2023-68318.1

Page 5 of 15