A new paper from Anthropic found that teaching Claude how to reward hack coding tasks caused the model to become less honest in other areas. 

Despite years of effort to make software safer and more transparent with SBOMs, the rise of AI coding assistants is fueling optimism—and, some experts argue, “kind of…

The spyware vendor made those two arguments, among others, in a motion to stay the California court ruling.

Visa CISO Subra Kumaraswamy joins Safe Mode to discuss the global scale and complexities of cybersecurity at Visa, from managing a billion transactions daily to maintaining a…

The letter comes as countries in Europe have moved over the past year to regulate or mandate legalized access for criminal and national security investigations.

Anthropic and AI security experts told CyberScoop that behind the hype, effective AI-driven cyberattacks still require skilled humans, with the attack possibly done to send a message…

Public Citizen’s letter urges OpenAI to temporarily take Sora 2 offline and work with outside experts to prevent the spread of harmful deepfakes.

Researchers warn that although exploitation of the zero-day is complex, a functional exploit exists in the wild.

Researchers say the nation-state attacker could cause more serious problems with the BIG-IP source code it nabbed during the attack on F5’s systems.

The security vendor said the attack, which exposed customers’ firewall configuration files, is contained and unrelated to recent Akira ransomware attacks on its customers.