#multi-turn attacks

Cisco Uncovers Major Weaknesses in Leading AI Models

Relying solely on single-turn benchmarks could mislead your AI security assessments.