1 articletagged with “TAR”
Adversarial attacks on AI-powered e-discovery systems: document classification manipulation, privilege prediction bypass, technology-assisted review poisoning, and predictive coding exploitation.