AIHighTrending
Apply Old Security Tactics to New AI Threats
Security experts are adapting traditional red teaming methods to find flaws in generative AI. This helps companies use frameworks like MITRE ATLAS to protect AI models from new threats like data poisoning and model hijacking before deployment.
InfoQ3 min read