Scientists developed an AI monitoring agent to detect and stop harmful outputs November 20, 2023 by DIGIT Spread the love The monitoring system is designed to detect and thwart both prompt injection attacks and edge-case threats. Go to Source
