Code Security | Breaking Cybersecurity News | The Hacker News

Ask an AI coding agent to scan open-source code for security holes, and it might run the attacker's code on your own machine instead. That is the finding in a  proof-of-concept published Wednesday by the AI Now Institute, an attack it calls " Friendly Fire. " It works against Anthropic's Claude Code and OpenAI's Codex when either is running in an autonomous mode that approves its own commands. It hijacks the exact job these tools are sold for: checking untrusted third-party code for problems. Instead of catching the threat, the agent becomes the way in. Researchers Boyan Milanov and Heidy Khlaaf tested two setups, each a stock install with the autonomous mode switched on: Claude Code (CLI 2.1.116, 2.1.196, 2.1.198, 2.1.199) on Claude Sonnet 4.6, Sonnet 5, or Opus 4.8 OpenAI Codex (CLI 0.142.4) on GPT-5.5 Claude Code's "auto-mode" and Codex's "auto-review" use a classifier to run commands the agent judges safe, pausing ...

An AI coding assistant that refuses to answer a dangerous request in its chat box can answer it anyway if the same request is broken into small, ordinary-looking steps inside a code editor. That is the finding of a  new study of GitHub Copilot  by researchers Abhishek Kumar and Carsten Maple. The models they tested through Copilot, Claude from Anthropic, and Gemini from Google, refused almost every harmful request when asked directly. Reframed as steps in a normal coding task, they produced the harmful answers in all 816 of the study's workflow runs. What makes this different from a typical jailbreak: no one asks for the harmful thing directly, and the model is not tricked into running someone else's code. It writes the banned content itself, as a side effect of a coding task it was told to improve. How it works The researchers call the method workflow-level jailbreak construction . Instead of a single blunt prompt, they asked Copilot to build an everyday piece of s...

Scanners meant to catch malicious add-on "skills" for AI coding agents can be fooled by a few simple changes that leave the malware working, according to a  new study  from researchers at the Hong Kong University of Science and Technology. Their strongest trick slipped past every scanner tested more than 90% of the time, and the same team built a runtime checker that catches most of the disguised skills the scanners miss. Skills are small packages, usually a Markdown instruction file plus a few scripts, that agents such as Claude Code, OpenAI Codex, and OpenClaw load to pick up a new capability. Because a skill is just a bundle of files, the same one can run across different agents. And it runs with the agent's own access: your files, your terminal, your saved passwords. A bad one can steal credentials, copy source code, or install a backdoor. Most of what a public marketplace lists is uploaded by strangers with little vetting. The main defense so far has been th...

OpenAI has announced the launch of an "agentic security researcher" that's powered by its GPT-5 large language model (LLM) and is programmed to emulate a human expert capable of scanning, understanding, and patching code. Called Aardvark , the artificial intelligence (AI) company said the autonomous agent is designed to help developers and security teams flag and fix security vulnerabilities at scale. It's currently available in private beta. "Aardvark continuously analyzes source code repositories to identify vulnerabilities, assess exploitability, prioritize severity, and propose targeted patches," OpenAI noted . It works by embedding itself into the software development pipeline, monitoring commits and changes to codebases, detecting security issues and how they might be exploited, and proposing fixes to address them using LLM-based reasoning and tool-use. Powering the agent is GPT‑5 , which OpenAI introduced in August 2025. The company describes it...

Eclipse Foundation, which maintains the open-source Open VSX project, said it has taken steps to revoke a small number of tokens that were leaked within Visual Studio Code (VS Code) extensions published in the marketplace. The action comes following a report from cloud security company Wiz earlier this month, which found several extensions from both Microsoft's VS Code Marketplace and Open VSX to have inadvertently exposed their access tokens within public repositories, potentially allowing bad actors to seize control and distribute malware, effectively poisoning the extension supply chain. "Upon investigation, we confirmed that a small number of tokens had been leaked and could potentially be abused to publish or modify extensions," Mikaël Barbero, head of security at the Eclipse Foundation, said in a statement. "These exposures were caused by developer mistakes, not a compromise of the Open VSX infrastructure." Open VSX said it has also introduced a toke...

A little over three dozen security vulnerabilities have been disclosed in various open-source artificial intelligence (AI) and machine learning (ML) models, some of which could lead to remote code execution and information theft. The flaws, identified in tools like ChuanhuChatGPT, Lunary, and LocalAI, have been reported as part of Protect AI's Huntr bug bounty platform. The most severe of the flaws are two shortcomings impacting Lunary, a production toolkit for large language models (LLMs) - CVE-2024-7474 (CVSS score: 9.1) - An Insecure Direct Object Reference (IDOR) vulnerability that could allow an authenticated user to view or delete external users, resulting in unauthorized data access and potential data loss CVE-2024-7475 (CVSS score: 9.1) - An improper access control vulnerability that allows an attacker to update the SAML configuration, thereby making it possible to log in as an unauthorized user and access sensitive information Also discovered in Lunary is anot...