Procurement
AI Vendor Subprocessor Chains and Third-Party Exposure
Enterprises inheriting hidden vendor chains when deploying AI-powered SaaS.
Tomáš Červenka
· 10 min read
Regulated Data Risk — Page 2
Samsung Confidential Source Code ChatGPT Leak Post-Mortem
Technical controls, not training alone, stop sensitive data from reaching AI services.
Mirela Vosburgh
Whistleblower and Journalist Data Exposure via AI Tools
Public AI tools store sensitive source material where courts can compel its disclosure.
Tomáš Červenka
ISO 42001 Audit Criteria for Privacy Preserving AI
Architecture choices like on-device inference prove privacy impact, not just good intentions.
Darnell Okafor-Baptiste
Retrieval-Augmented Generation and Cross-Tenant Data Leakage
RAG systems leak data across tenants by design, not accident.
Tomáš Červenka
AI Data Processing Addendum Key Clauses
Model training on customer data requires explicit authorization, not silence in your contract.
Darnell Okafor-Baptiste
AI Tool Exposure of Government Classified or Sensitive Data
Commercial AI tools retain user data by design.
Tomáš Červenka
SOC 2 Type II Reports for AI Services — What They Miss
SOC 2 Type II reports miss critical AI-specific risks like model drift and agentic behavior.
Darnell Okafor-Baptiste