Affiliate Ram Shankar Siva Kumar and coauthors suggest ways that Microsoft and other tech giants can mitigate the security risks inherent in emerging AI technologies.

OpenAI used its own o1-preview and o1-mini models to test whether additional inference time compute protected against various attacks.

American business services giant and government contractor Conduent confirmed today that a recent outage resulted from what it described as a "cyber security incident." South Korean VPN provider ...

Seven vulnerabilities, including BootHole and PixieFail, in Palo Alto firewalls risk Secure Boot integrity and enable sophisticated attacks.

SentinelOne researchers highlighted similarities in the approaches used by the HellCat and Morpheus ransomware groups, ...