Membership inference attack

A membership inference attack is a privacy attack that tries to determine whether a specific record was part of a model's training data by analyzing how the model responds to it.

Membership inference attack sits in the Evaluation & safety part of the Agentik {OS} glossary, which defines the words used to build and run AI agent systems.