An experimental OpenAI agent was given a routine research task, but when it hit a barrier, it found another path and gained unauthorized access to Australian government systems.
AI agents can access sensitive data, credentials, APIs, and internal systems at machine speed. In this sponsored Q&A, Dr. Gaurav Bhat, SVP of Symmatrics, explains the security risks companies need to consider before giving autonomous agents too much trust.
AI tools are helping researchers uncover Linux vulnerabilities at a faster pace. Canonical is responding by changing Ubuntu’s kernel release strategy to get security fixes into users’ hands sooner.
The Linux Foundation is bringing companies and researchers together to build an open defensive stack for securing AI models, agents, and infrastructure.
Surfshark says human error exposed an internal test server to the internet, allowing an unauthorized party to access engineering material. The VPN provider says customer data and production systems were not affected.
Microsoft is bringing Codename MDASH to Azure Government, unleashing more than 100 specialized AI agents to hunt vulnerabilities in government source code and determine which flaws could actually be exploited.
Hackers are actively exploiting the StyleSmuggler zero-day in Magento and Adobe Commerce, using it to compromise online stores and deploy malware on Linux servers.
Surfshark has opened an updated beta testing program that lets Windows and Android users try unfinished features and report bugs before public releases.
Cloudflare is combining its live Internet security data with OpenAI GPT-5.6 Cyber to prioritize vulnerabilities, block active attacks at the edge, and help developers generate fixes.
Palo Alto Networks has acquired AI-native platform Console as it pushes Cortex toward security agents that can investigate threats and take action.
Cloudflare Adaptive Intelligence continuously retrains its bot detection system and will eventually generate disposable rules designed to keep attackers from learning how to defeat its defenses.
AI is accelerating software attacks while security teams struggle to patch vulnerabilities. New Contrast Security research shows just how wide that gap has become.