The AI you didn't build: From black box to defensible risk

Brought to you by TrustArc

Most AI risk today isn't homegrown, it's embedded in the third-party software you already use. When privacy and risk teams can't inspect the model, "high risk" isn't a defensible answer.

A practical framework is needed for discovering AI hidden in vendor products, mapping it to data and business processes, and scoring risk in a way that holds up to leadership, auditors or regulators. The outcome: separating inherent from residual risk, translating findings into procurement and contract terms, and building the evidence trail to answer, "How did you arrive at that number?"

• View presentation