AI is changing how security teams find vulnerabilities, analyze code, test applications, and protect infrastructure. Developers are building tools to secure AI systems themselves, from coding agents and memory protection to model exposure discovery. This roundup covers recent open-source releases…
Category: Help Net Security
macOS is becoming a proving ground for AI agents
Somewhere right now, a Mac Mini is sitting on a shelf doing someone’s chores. Nobody’s watching it. It reads a version number out of Terminal, hops over to Safari, digs up a release year, then quietly files a reminder, the…
How to implement a continuous offensive security testing program
The hard part was never finding the exposure. It was deciding what to do about it: whether to patch, mitigate, monitor, or accept, and banking that that decision would still hold tomorrow. A penetration test answers this question for the…
OpenAI and Anthropic are pulling in different directions
Companies are handing routine operational decisions to AI agents that plan, remember, and act on their behalf. These agents run on statistical models, and their behavior can drift across weeks and months. That drift opens a security gap outside the…
Picus Autonomous Exposure Validation Platform validates real-world CVE exploitability
Picus Security has launched the Picus Autonomous Exposure Validation Platform, built for a world where frontier AI has collapsed the time between disclosure and attack. Adversaries now weaponize new CVEs in hours, against a backdrop of around 132 published every…
CyberProof Agentic MXDR Service brings AI agents to managed detection and response
CyberProof has announced the launch of the CyberProof Agentic MXDR Service which connects AI agents with human expertise and presents quantifiable security outcomes with CyberProof’s Reveal360. CyberProof modernizes managed detection and response by shifting security operations from manual workflows to…
Barracuda adds PAM and identity protection with Evo Security acquisition
Barracuda Networks has acquired Evo Security. The acquisition expands the BarracudaONE platform’s identity security capabilities by adding privileged access management (PAM), access control, identity protection, and identity threat detection and response. By combining Evo Security’s identity solutions with Barracuda’s existing…
Commvault measures cyber recovery readiness with AI attack simulations
Commvault has announced Commvault Minutes to Recovery, a scenario-driven cyber resilience simulation that lets participants act as a hacker and run their own attacks using frontier AI tools. Then, participants are challenged to defend against and recover from an incident…
Attackers exploit critical Adobe ColdFusion vulnerability (CVE-2026-48282)
CVE-2026-48282, one of the maximum severity vulnerabilities patched in Adobe ColdFusion on June 30, 2026, has been targeted by attackers in the wild. Exploitation attempts were detected on July 2, through the honeypot sensors of cybersecurity threat-intelligence service KEVIntel, mere…
Radware updates Agentic AI Protection with AI governance and compliance capabilities
Radware has announced enhancements to its Agentic AI Protection solution to help organizations govern and secure AI agents across enterprise environments. The release adds compliance reporting to support alignment with leading global AI standards, enhanced visibility into agent ecosystems, and…
Researchers make the case for a cybersecurity AI scientist
Autonomous AI agents have started doing real security work. Language-model agents probe software for flaws, run penetration tests, and chain together attack steps that once needed a human operator. Research about security has stayed slower and more manual, built around…
Review: Building Machine Learning Systems with a Feature Store
Many people come to machine learning by training a model on a tidy dataset, and then meet a harder problem: making that model work for real users, on fresh data, every day. Jim Dowling’s O’Reilly book, Building Machine Learning Systems…
Your company already adopted AI and nobody is governing access
In this Help Net Security video, Antoine Berton, CTO at Elba Security, breaks down the AI attack surface. Your company already adopted AI, and every adoption creates access that nobody governs. A quick click on a Friday afternoon connects a…
Apple Container: Open-source tool for Linux containers on the Mac
Developers on Apple silicon Macs have run Linux containers through software built around a single shared virtual machine for years. Apple’s open-source Container project gives each Linux workload its own lightweight virtual machine. Container is written in Swift and tuned…
Power shortages could slow AI data center expansion
AI adoption is increasing demand for data center capacity at the same time operators are running into limits around power, equipment, land, and permitting, according to NTT Data. Access to electricity is becoming a deciding factor in where new data…
Microsoft wants to keep your AI agents from going rogue
Microsoft has introduced Microsoft Execution Containers (MXC), a cross-platform, policy-driven execution layer for AI agents on Windows and Windows Subsystem for Linux (WSL), now available in early preview. Updated Agent 365 platform (Source: Microsoft) Developers can define constraints for their…
Cybersecurity jobs available right now: July 7, 2026
Application Security Lead Gett | Israel | Hybrid – View job details As an Application Security Lead, you will lead application and cloud security initiatives by integrating security into the SDLC, overseeing threat modeling, secure architecture, application security testing, and…
Criminal IP integrates threat intelligence with OpenCTI for automated indicator enrichment
Criminal IP has integrated its threat intelligence with OpenCTI, enabling security teams to automatically convert IP addresses, domains, and URLs into structured intelligence within the platform’s knowledge graph. The integration automatically enriches ingested indicators with Criminal IP’s infrastructure intelligence, dual-perspective…
OpenSSH 10.4 arrives with security fixes and a post-quantum signature option
Operators who manage remote access to Unix and Linux systems keep a close watch on OpenSSH, the software that carries most SSH traffic across the internet. The project released version 10.4 with eight security fixes, a set of bug corrections,…
LTM’s BlueVerse RightLogic combines AI risk assessment with cyber remediation planning
LTM has launched BlueVerse RightLogic, a cybersecurity assessment and risk assurance framework designed to help enterprises identify, assess and remediate cyber exposure as they accelerate AI adoption. AI is now capable of autonomously identifying and exploiting vulnerabilities, while exposure across…