ActiveFence Releases AI Security Benchmark: Industry-Leading Precision and F1 in Detecting Prompt Injections

ActiveFence, the AI Safety company protecting enterprises from GenAI misuse and misalignment, today announced the publication of its AI Security Benchmark Report: Prompt Injections. The study evaluates six leading guardrails and APIs on their ability to detect adversarial prompt attacks, showing ActiveFence's AI Safety & Security model with the top F1 score (0.857) and precision (0.890) while maintaining a competitive false-positive rate (5.4%).