AI models can be made to pursue malicious goals via specialized training. Teaching AI models about reward hacking can lead to other bad actions. A deeper problem may be the issue of AI personas.

Anthropic CEO Dario Amodei announces the company’s new models at the Code with Claude conference in San Francisco, May 22, 2025 (photo by Don Feria/AP Content Services for Anthropic). Frontier AI ...

Plus: State-sponsored AI hacking is here, Google hosts a CBP face recognition app, and more of the week’s top security news. The United States issued a seizure warrant to Starlink this week related to ...

The company claimed that A.I. did most of the hacking with limited human input and said it was a rapid escalation of the technology’s use in cybercrime. By Meaghan Tobin and Cade Metz Meaghan Tobin ...

Crowdsourced cyber security firm Bugcrowd hopes to make good on a plan to “unite the hacker community and the power of AI” after acquiring Mayhem Security, an artificial intelligence (AI) and cyber ...

The Department of Justice (DOJ) is facing pressure to back away from a request from President Trump for a $230 million settlement stemming from his legal troubles, as critics say it raises a dizzying ...

Linux I'm brave enough to say it: Linux is good now, and if you want to feel like you actually own your PC, make 2026 the year of Linux on (your) desktop AI 'There is *zero* point in talking about AI ...

Have you ever wondered how secure your Wi-Fi network really is? With cyber threats becoming increasingly sophisticated, even the most robust-looking networks can harbor vulnerabilities. But what if ...

AI agents are automating key parts of the attack chain, threatening to tip the scales completely in favor of cyber attackers unless new models of AI-assisted cyberdefense arise. AI agents are now ...

In the last 30 years or so, cybersecurity has gone from being a niche specialty within the larger field of computer science, to an industry estimated to be worth more than $170 billion made of a globe ...