General cybersecurity news and analysis
A security flaw in the Claude AI Chrome extension could put users at risk, as it allows other browser extensions to issue commands to Claude without verification. This vulnerability creates a backdoor for hackers to hijack the AI model, warns LayerX senior researcher Aviad Gispan.
Senate Minority Leader Chuck Schumer is sounding the alarm on the urgent need for stronger AI cyber defenses, warning that there's a high-stakes race between cybersecurity defenders and AI-enabled hackers. He’s pressing the Department of Homeland Security to team up with state and local governments to stay ahead of rapidly evolving threats.
Researchers uncovered a shocking 91% of autonomous AI agent deployments are vulnerable to tool-chaining attacks, revealing a critical weakness in current governance approaches. This startling finding highlights the urgent need for updated security measures to protect AI systems in healthcare, finance, customer service, and software development.
The harsh reality is that security operations centers (SOCs) are drowning in a sea of alerts, with a typical workload of 120-150 alerts per day, which translates to 40-50 analyst-hours of work - far exceeding the capacity of most teams. This means many alerts are left uninvestigated or pushed to the next shift, leaving SOCs vulnerable to threats.
Meta has backtracked on its plan to introduce end-to-end encryption for Instagram messages, leaving direct messages stored and transmitted in plaintext, giving the company access to user conversations. This move reverses years of promotion for enhanced messaging security.
The US government is overhauling its CyberCorps Scholarship For Service program, rebranding it as CyberAI SFS, to prioritize AI skills in cybersecurity. Students enrolled in the program will now need to be proficient in using AI to stay employable after graduation.
A critical vulnerability, CVE-2025-68670, was discovered in the xrdp remote desktop server, allowing for remote code execution - a flaw that was thankfully patched in January 2026. This security risk was found during a routine audit, highlighting the importance of regular security checks to protect against potential threats.
Don't let low-severity alerts fly under the radar - nearly 1% of confirmed incidents come from these seemingly minor warnings, translating to around one missed breach per week for a typical enterprise. This small but significant gap in enterprise security can have big consequences.
Evidence in spreadsheets, controls drifting between audits, frameworks multiplying on flat headcount. Nubivance runs continuous compliance on Rapid7 Cyber GRC - SOC 2, HIPAA, ISO 27001, PCI, CMMC.
End the scramble
A contractor's reckless actions led to the destruction of dozens of federal databases, showcasing a staggering disregard for the security and integrity of sensitive government information. After being terminated on February 18, 2025, the contractor and his twin brother intentionally caused chaos by accessing computers without authorization and deleting crucial data.
The Pentagon is shaking up its AI strategy, ditching the single-vendor approach and embracing a multi-provider model to ensure seamless integration with classified systems. This bold move follows a hard-won lesson: relying on one AI vendor just won't cut it.
The Pentagon is banking on AI tools to supercharge its cyber defense capabilities, with Assistant Secretary for Cyber Policy Katherine Sutton touting their potential to develop secure code in minutes to seconds. This revolutionary speed could transform the current defensive posture, outpacing threats and safeguarding against vulnerabilities at unprecedented rates.
Mozilla's April bug cull was massive, with 423 Firefox security fixes - a whopping five times more than the previous month and 20 times the usual monthly average, thanks in part to a boost from AI-powered bug detection. This huge spike in repairs is a testament to the power of innovative technology in keeping Firefox users safe and secure.
WatchGuard's acquisition of Perimeters is a strategic move to supercharge its cloud security offerings, driven by a personal endorsement from its CISO who fell in love with Perimeters' technology. This exciting purchase brings Perimeters' innovative cloud application security solutions into WatchGuard's portfolio.
Microsoft Edge's password management has a concerning vulnerability: it loads all saved passwords into browser memory in plaintext at startup, making it easier for hackers to steal credentials on compromised systems. This is in stark contrast to other Chromium-based browsers like Google Chrome and Brave, which only decrypt passwords when needed.
A single click on Claude Code's generic dialog can unleash a major security risk, allowing an unsandboxed Node.js process to spawn with full user privileges. This vulnerability can be exploited using just two common JSON files, putting developers at risk of one-click remote code execution.
As organisations navigate the complexities of multi-cloud estates, modern cloud firewalls are emerging as a crucial linchpin for reclaiming coherent security controls and mastering cloud networking. They're not old tech, but a vital tool for cloud architects and security pros to enforce robust, zero-trust security across multiple providers.
Full-platform Rapid7 rollouts - InsightVM, InsightIDR, event sources, agents, tuning - by a Registered Partner who has done it at state-agency scale, including OT/SCADA.
Talk deployment
When a cyber breach hits, the decisions made in the first few days can have a lasting impact, setting the stage for years of consequences - and it's not just about fixing the tech, but also about the legal and communication choices that are made early on. In fact, a single incident can generate a ripple effect of legal, regulatory, and reputational consequences that persist for years.
In a shocking test, Kaspersky researchers cracked 60% of 231 million MD5 password hashes in under an hour using just one high-powered graphics card, revealing the alarming vulnerability of even the most seemingly secure passwords. This unsettling experiment highlights the urgent need for stronger password protection.
Bitcoin Core developers have disclosed a high-severity vulnerability, tracked as CVE-2024-52911, which is the project's first known memory safety flaw that could potentially allow remote code execution. This rare but critical bug was fixed months ago and affects Bitcoin Core releases from 2017 to early 2025.
Storing passwords in plain text poses a significant risk, especially in shared environments, as a security researcher recently discovered that Microsoft Edge saves decrypted credentials in its memory, making them vulnerable to exposure. This flaw allows saved passwords to be accessible even when they're not in use.
On World Password Day, cybersecurity experts are sounding the alarm: it's time to rethink our reliance on passwords, as attackers continue to exploit weak visibility and poor credential management to gain access to sensitive systems. The real vulnerability isn't a single weak password, but how credentials spread across organizations, often with employees reusing and sharing access without centralized tracking.
A critical vulnerability in Cline Kanban's WebSocket endpoints lets hackers hijack websites visited by developers, silently interacting with local AI agents - and it's a flaw that requires zero phishing, malware, or social engineering. This severe flaw, scoring 9.7 on the CVSS scale, puts AI coding agents at risk of website hijacking.
With data constantly moving across cloud and AI environments, traditional security tools are holding you back from truly protecting your data - it's time for a modern approach. A staggering 72% of security professionals agree that data security is more critical than ever, making an evolution in strategy urgent.
Traditional data loss prevention methods are struggling to keep up as 46% of sensitive file uploads to web apps are sent to unsanctioned accounts, revealing a significant blind spot in modern DLP systems. This gap is largely due to the rise of browser-based activities, where conventional DLP methods fall short.