OpenAI has taken a more aggressive approach to red teaming than its AI competitors, demonstrating its security teams' advanced capabilities in two areas: multi-step reinforcement and external red ...

Red teaming is a powerful way to uncover critical security gaps by simulating real-world adversary behaviors. However, in practice, traditional red team engagements are hard to scale. Usually relying ...

Nearly every organization today works with digital data—including sensitive personal data—and with hackers’ tactics becoming more numerous and complex, ensuring your cybersecurity defenses are as ...

Unrelenting, persistent attacks on frontier models make them fail, with the patterns of failure varying by model and developer. Red teaming shows that it’s not the sophisticated, complex attacks that ...

AI red teaming now has a government-defined standard for the first time worldwide. South Korea’s Ministry of Science and ICT ...

Microsoft says it needs to strengthen trust in AI, which cannot be achieved without a full understanding of the changes AI generates. In Microsoft’s AI Red Team, alongside cybersecurity experts, ...

As generative AI transforms business, security experts are adapting hacking techniques to discover vulnerabilities in intelligent systems — from prompt injection to privilege escalation. AI systems ...

With plenty of pentesting tools out there you must know how they work and which fits the use case you are interested in testing. CSO selected 14 underrated tools and what they are best for. The right ...

Generative artificial intelligence (GenAI) has emerged as a significant change-maker, enabling teams to innovate faster, automate existing workflows, and rethink the way we go to work. Today, more ...

A tool for red-team operations called EDRSilencer has been observed in malicious incidents attempting to identify security tools and mute their alerts to management consoles. Researchers at ...

The group responsible for red teaming of over 100 generative AI products at Microsoft has concluded that the work of building safe and secure AI systems will never be complete. In a paper published ...