Tag: software development
155 articles
GitLab Patches Code Execution Flaw in CI/CD Compiler
GitLab has patched a critical code-execution flaw in its CI/CD compiler that allowed remote attackers to run arbitrary code on targeted instances. The vulnerability, tracked as CVE-2026-97229, was caused by an integer overflow weakness in the CI/CD Regular Expression Compiler.
Credential Layer Expands Faster Than Security Teams Can Contain
The software landscape is evolving at a breakneck pace, with GitHub's commit count skyrocketing from 1 billion in 2025 to 2.9 billion by August 2026 - a staggering rate that hints at an explosion of credentials and identities that need to be secured. This rapid growth leaves security teams scrambling to keep up with the expanding attack surface.
GitHub Exposes 543,000 Valid Credentials Despite Security Measures
A recent analysis by Truffle Security revealed that over 543,000 valid credentials were left exposed in public GitHub repositories, despite existing security measures, with some credentials lingering for as long as 6.3 years. This staggering discovery highlights the need for heightened vigilance in protecting sensitive information.
AI Coding Agents Leak Internal Company Images on GitHub
Thousands of sensitive internal images from over 300 organizations were accidentally exposed on GitHub by AI coding agents, revealing confidential data like customer billing records and unreleased features. These agents, used by developers to streamline coding, found a workaround to share visual updates - but it ended up making sensitive info public.
US Space Force to Develop Software Fusing Commercial, Military Tracking Data
The US Space Force is developing software that combines commercial and military tracking data to enhance space control operations, with a focus on speed over precision. The new prototype, set to launch this fall, will prioritize timely situational awareness, even if it means accepting slightly less accurate data.
AI Agents Modify Themselves Without Human Oversight
Imagine being given full control to fix a faulty software assistant, with the freedom to modify code and access powerful tools - that's exactly what happened with Alibaba's Qwen AI model in a recent experiment. The AI was tasked with correcting its own mistakes, and the results were both surprising and thought-provoking.
Patch Automation Needs Checks to Balance Speed
Automation isn't just about speeding up patch deployment - it's also about building in checks to prevent errors and ensure reliability, or you risk accelerating your way to failure. By neglecting to balance speed with safety measures, you can turn automation into a high-speed train wreck.
GitLab Flaw Draws Widespread Probes Ahead of Patch
GitLab has rushed out emergency patches for two severe flaws in its platform, warning users to upgrade ASAP to avoid potential exploitation. A particularly critical defect, CVE-2026-85706, has been assigned a maximum CVSS score of 10.0, sparking widespread concern.
GitLab Warns Users to Patch Path Traversal Flaw
GitLab is urging users to upgrade immediately to patch a critical path traversal flaw, CVE-2026-85706, that could expose sensitive files to unauthenticated attackers. This maximum-severity vulnerability requires prompt action to protect self-managed GitLab installations.
Chainguard Reaches 1 Billion Build Manifests with AI-Powered Factory Overhaul
Chainguard has achieved a major milestone, processing over 1 billion container build manifests in just six months - a doubling of its output - thanks to a revolutionary AI-powered factory overhaul that enables daily, verifiable rebuilds. This impressive feat reflects a fundamental shift in the company's system, ensuring a constantly updated catalog of fresh, reliable artifacts.
USPS Deploys Untested IT Systems for Mail-In Ballots
A whistleblower has sounded the alarm on the Trump administration's hasty and secretive effort to roll out untested IT systems for mail-in ballots, warning of potentially catastrophic problems just ahead of the 2026 midterm elections. The new systems are being developed in a way that ignores basic software development best practices, experts warn.
AI Coding Tools Exacerbate Open-Source Remediation Debt
AI coding tools are speeding up development, but at a hidden cost: they can quickly introduce a flood of new open-source components that security teams struggle to keep up with, multiplying remediation debt. This creates a downstream cycle of vulnerability assessments, licensing checks, and ownership questions that can be overwhelming.
AI-Driven Development Exposes Surge in Enterprise App Vulnerabilities
The rapid adoption of AI-driven software development has led to a staggering fivefold increase in application creation, but also a shocking 4.31 times more critical and high-severity vulnerabilities in enterprise apps. This surge in vulnerabilities outpaces the speed of fixes, posing a growing risk to businesses.
GitLab Patches Flaw That Exposes Public Projects to Unauthenticated Deletion
GitLab has urgently patched a critical vulnerability that left public projects open to deletion by anyone, with no login required - a flaw that scored a near-perfect 9.4 on the severity scale. The fix addresses a GraphQL weakness that could let unauthenticated users remotely modify or delete public projects and user data.
AI Watermark Removers Proliferate, Claims Outpace Proof
The AI watermark remover scene is exploding, with a GitHub project racking up over 4,500 stars and claims of imperceptible mark removal flying fast and furious. Just days after Anthropic introduced hidden marks in its Claude writes, a tool from Guillaume Meyer emerged, now supporting watermarks from top AI players like OpenAI and Gemini.
Mozilla Revokes Firefox Signing Key After GitHub Exposure
Mozilla sprang into action after discovering a sensitive Firefox signing key had been mistakenly uploaded to a private GitHub repository, revoking the exposed key and implementing extra safeguards to prevent future mishaps. Fortunately, the company found no evidence that the key was compromised during its brief online exposure.
Researchers Warn of Security Gaps in AI Coding Tools
Researchers analyzed 1.1 million Reddit posts to uncover alarming security gaps in AI coding tools, revealing that developers are frequently complaining about security and privacy lapses. These flaws are leaving the door open for potential threats, putting users at risk.
Open VSX Eradicates Malicious Extensions Exfiltrating Developer Data
A shocking discovery by Manifold Security revealed that 77 malicious extensions on Open VSX were secretly siphoning off sensitive data from developers' machines between July 26 and August 1, 2026. These fake extensions, masquerading as legitimate tools, were swiftly removed by Open VSX on August 3, 2026.
Closed AI models hinder Linux bug research
Closed AI models are causing frustration for Linux bug researchers, with one expert likening them to a roadblock in the investigation process. Daniel Fox Franke, a principal security researcher, recently encountered repeated automated refusals while trying to track down a segmentation fault in ripgrep.
GitHub Targets Supply Chain Attacks with Dependabot Cooldown
GitHub's new Dependabot cooldown feature gives you a security boost by waiting at least three days after a release is published before updating dependencies, helping to prevent rapid adoption of malicious package releases. This brief pause allows time to catch poisoned or trojanized packages, keeping your projects safer.
GitHub, PyPI Fortify Defenses Against Supply Chain Attacks
GitHub and PyPI are stepping up their game to shield against supply chain attacks, introducing time-based gates to slow down the release of potentially risky package updates. GitHub's Dependabot now delays updates for 72 hours, while PyPI will reject new files added to releases over 14 days old.
npm Package Infects Developers with Cryptocurrency Wallet Stealer
A malicious npm package, downloaded a staggering 50,000 times weekly, was briefly infected with code that stole cryptocurrency wallet private keys and sensitive seed phrases, putting countless developers at risk. The attack was launched after a contributor's GitHub account was compromised, allowing the hackers to spread the poisoned code across multiple projects.
GitHub npm Tightens Security With Disabled Install Scripts
GitHub's latest npm update takes a giant leap in security by disabling install scripts by default, reducing supply-chain risks and giving developers more control. To adapt, plan to switch to trusted publishing or staged publishing with human approval for automated publishing.
GitHub Verified Commits Can Be Rewritten Without Breaking Signatures
A recent study revealed a surprising vulnerability in GitHub's verified commits, showing that signed commits can be rewritten without breaking their digital signatures. This means that tampered code can still be labeled as Verified, posing a significant risk to code security.