AI Security | Breaking Cybersecurity News | The Hacker News

Everyone seems to have announced a clearinghouse over the past few weeks. We did too. Ours is called Athena , and the main thing that sets it apart is that it was already real and running when we announced it — built quietly months earlier, heads down, taking findings and shipping fixes, because customers kept asking us to. We only announced it now because everyone else started announcing theirs, and staying quiet started to look like something it wasn't. The others arrived louder and, as far as anyone outside the press releases could tell, didn't exist yet. Here's the part none of those announcements will tell you: the clearinghouse is the least important thing to build. When a project we'd deliberately kept private, a  five-billion-dollar press release , and  the White House all reach for the same word inside a few weeks, that's not a trend. Trends are optional. This is the shape of a problem changing under everyone at once. So let me explain why these thin...

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 ...

Researchers at  Wiz  found that a flaw in six popular AI coding assistants lets a booby-trapped code project quietly take control of a developer's computer. The assistant asks permission to edit one harmless-looking file, but the write lands on a sensitive one instead. The affected tools are Amazon Q Developer, Anthropic's Claude Code, Augment, Cursor, Google Antigravity, and Windsurf. Wiz calls the pattern GhostApproval and published it on July 8. Three of the six have shipped fixes, two have not, and Anthropic disputes that it is a bug. The most exposed are the tools that change files before you can weigh in. How the attack works The attack abuses an old Unix feature called a symbolic link , or symlink , that the assistants fail to check. A symlink quietly points to another file elsewhere on disk, so writing to it actually writes to the target. Wiz built a malicious repository with a symlink named project_settings.json that really points to the victim's ...

Sophos looked at a week of its own endpoint data and found that AI coding agents such as Claude Code, Cursor, and OpenAI Codex are setting off detection rules written to catch human intruders. The agents are not malicious. They just do a lot of things that, to a behavioral engine, look exactly like an attack. Decrypting browser credentials, listing what sits in Windows' credential store, pulling files down with built-in system tools, writing to the startup folder: these have long been high-signal to defenders. What has changed is who is generating it. On the machines Sophos watched, it was often a developer's AI assistant going about ordinary work. What set the alarms off The  analysis  draws on seven days of telemetry from June 2026, taken from Sophos's behavioral engine on Windows and counted by unique machines, not raw event volume. It is a narrow window on one vendor's fleet, not an industry census. Sophos's charts put credential access at 56.2 perc...

AI coding assistants have a habit of making things up. Ask one to fetch a popular tool, and it will sometimes hand back a real-sounding name for a project that does not exist. New research, which its authors call  HalluSquatting , turns that habit into an attack: work out the fake names an AI reliably invents, register them first, and wait for the assistant to fetch your trap on a user's behalf. Anyone whose AI assistant can fetch an outside resource and then run commands with little human review is exposed. In tests, that path led the assistant to run attacker-supplied code on the machine. Repeat it with a popular enough resource, and one planted name can reach many machines, which is why the researchers frame it as a way to assemble a botnet. How it works The attack chains two AI quirks. The first is a  hallucination : an AI making something up and presenting it as real. The second is a  prompt injection : a booby-trapped instruction that hijacks the AI, so i...

A new banking fraudulent operation is targeting customers of Mexican banks, fintech, payment processors, and cryptocurrency exchanges using ClickFix lures. The activity cluster, tracked by Elastic Security Labs under the moniker REF6045 , involves infecting victims through fake CAPTCHA verification pages that deceive them into running a malicious command that installs a PowerShell toolkit dubbed SCMBANKER . Some components of the malware date back to October 2025. "Once installed, the operator can see when a victim opens a banking session, lock the screen behind a fake bank warning, push the victims towards live phone interaction, redirect the browser, or replace account numbers copied to the clipboard," security researchers Jia Yu Chan and Salim Bitam said . "For a full takeover, they can also deploy a commercial remote-access tool." SCMBANKER is specifically designed to go after Mexico's financial ecosystem, with evidence pointing to the use of a large...

For years, account takeover (ATO) followed a predictable script. Attackers bought stolen credentials in bulk, ran them through automated tools, and waited for matches. Credential stuffing was cheap, scalable, and for defenders, relatively well understood. That era is ending. Not because attackers gave up, but because the front door finally got harder to kick in. Passkeys are now mainstream. According to the FIDO Alliance's 2026 research, 75% of global consumers have enabled a passkey on at least one account. At the same time, passkeys are becoming more common in the workplace, with 68% of companies now using, testing, or introducing them for employee sign-ins.  Phishing-resistant, passwordless authentication is no longer aspirational, it's becoming the default. When the password disappears, so does the value of a stolen password. So where does the attack go next? It moves downstream, to the moments where systems still trust a human to prove who they are. The attac...

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...

The U.S. Cybersecurity and Infrastructure Security Agency (CISA) on Tuesday added four security flaws to its Known Exploited Vulnerabilities ( KEV ) catalog, citing evidence of active exploitation. The vulnerabilities are listed below - CVE-2026-48282 (CVSS score: 10.0) - A path traversal vulnerability in Adobe ColdFusion that could lead to arbitrary code execution in the context of the current user. CVE-2026-56290 (CVSS score: 10.0) - An improper access control vulnerability in Joomlack Page Builder that could allow for remote code execution via unauthenticated arbitrary file upload. CVE-2026-55255 (CVSS score: 6.1) - An authorization bypass through a user-controlled key vulnerability in Langflow that could allow an authenticated attacker to execute any flow belonging to another user by specifying the victim's flow ID in the request. CVE-2026-48908 (CVSS score: 10.0) - An unrestricted upload of a file with a dangerous type vulnerability in JoomShaper SP P...

A critical flaw in Google's Dialogflow CX could have let an attacker with edit rights on one Code Block-enabled agent compromise other Code Block-enabled agents in the same Google Cloud project. From there, they could read live conversations, steal the data users shared, and make the bots send attacker-written messages, including requests to re-enter a password. Security firm Varonis found it and named it Rogue Agent. The flaw affected only organizations that built agents with Dialogflow's Playbooks and custom Code Blocks, which let developers add their own Python. And it was not a remote, unauthenticated attack. Pulling it off needed the dialogflow.playbooks.update permission on one such agent, which limits the realistic attacker to a malicious insider or a compromised developer account, not a stranger on the internet. From that one foothold, though, the reach extended to every agent in the project. Google has fixed it, and both Varonis and Google say there is no sig...

A public issue can trick GitHub Agentic Workflows into leaking the contents of an organization's private repositories, researchers at Noma Security have shown. The attacker needs only to open a normal-looking issue on a public repository, with no stolen credentials and no access to the organization. If that organization has given the agent read access across its repositories, private ones included, the issue can steer it into pulling private contents into a public comment. Noma calls the technique GitLost . The target is GitHub Agentic Workflows , a feature now in public preview that GitHub launched in February. Instead of writing automation scripts, you write instructions to an AI agent in plain English in a Markdown file. The agent reads issues and pull requests, runs tools, and replies on its own. It can be powered by GitHub Copilot, Anthropic's Claude, Google Gemini, or OpenAI Codex. Workflows are read-only by default, but an organization can hand one a token with...

Cybersecurity researchers have disclosed details of a now-patched critical session isolation vulnerability in Writer , an enterprise generative artificial intelligence (AI) platform, that could result in cross-tenant compromise. The one-click vulnerability has been codenamed WriteOut by the Sand Security Research team. "An outsider could go from having no access to taking over any Writer AI organization inside industry-leading enterprises, with nothing more than a link," the cybersecurity company said in a report shared with The Hacker News. Put differently, the shortcoming could be abused to take over a victim's Writer account, and use it to access private chats, documents, and other sensitive data related to agents, configurations, private models, connectors, and large language model (LLM) credentials. Even worse, it could be abused to seize administrative control depending on the victim's role. An important aspect of the flaw is that the attacker and the...

Software supply chain security was hard enough. Then AI joined the build pipeline. For five years, "software supply chain security" meant one question: what's in your code? Which open-source packages, which versions, which transitive dependencies three layers deep that nobody chose on purpose? SolarWinds, Log4Shell, and XZ Utils all taught the same lesson: the risk lives less in the code a team writes and more in everything that produces it. Shai-Hulud, the self-propagating malicious package campaign that spread through developer toolchains this year, taught the next one: knowing what's in your code is still necessary, but it's no longer sufficient. In the roughly 20 months since the Model Context Protocol launched, AI tools, models, and the infrastructure around them have become load-bearing parts of how software gets built, deployed, and run. Code is written by agents. Packages are pulled in by autonomous tools that decide they are needed. Prompts have...

BeyondTrust has released updates to address two critical security flaws affecting Remote Support (RS) and Privileged Remote Access (PRA) products that, if successfully exploited, could allow unauthenticated attackers to take control of susceptible devices. The vulnerabilities are listed below - CVE-2026-40138 (CVSS score: 9.2) - A pre-authentication vulnerability exists in the authentication subsystem of BeyondTrust Remote Support and Privileged Remote Access stemming from improper validation of authentication data that could allow a network-positioned attacker to bypass access controls and gain unauthorized access to the appliance, including accounts with elevated privileges. CVE-2026-40139 (CVSS score: 9.2) - A pre-authentication vulnerability exists in the authentication subsystem of BeyondTrust Remote Support stemming from improper processing of authentication requests that could allow an unauthenticated remote attacker to bypass access controls and gain unauthorized ...

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...

The safety check that is supposed to stop an AI coding agent from running a dangerous command can be walked straight past using a shell trick that has been public for decades. New research from  Adversa AI , which is named the bypass GuardFall , found it works against ten of the eleven popular open-source coding and computer-use agents the firm tested. Only one, "Continue," was built to defend against it. Why does it matter? These agents run shell commands with your full account access. Point one at a booby-trapped repository or software package, and a hidden instruction can quietly run a command that wipes files or steals the secrets your account can reach, from SSH keys and cloud credentials to anything sitting in your home folder. How does it get past the guard? Most of these agents try to stay safe by checking each command against a blocklist of dangerous patterns before running it. The flaw is that they check the command as plain text, while bash rewrites that t...

An unknown threat actor has been observed exploiting a recently disclosed maximum-severity security flaw in SimpleHelp to deliver two previously unreported malware families, TaskWeaver and Djinn Stealer . The intrusion involves the exploitation of CVE-2026-48558 (CVSS score: 10.0), a critical authentication bypass vulnerability impacting the OpenID Connect (OIDC) flow that an unauthenticated attacker could exploit to obtain a fully authenticated "Technician session by submitting a forged token containing arbitrary identity claims. "TaskWeaver is a heavily obfuscated Node.js loader, delivered as jquery.js and executed through node.exe, that implements an encrypted, reusable payload delivery channel rather than a fixed set of post exploitation commands," Blackpoint Cyber said in an analysis. "The observed second stage payload, Djinn Stealer, targets Windows, macOS, and Linux systems." Djinn Stealer is designed to harvest credentials associated with cloud...

Microsoft has found a malicious Chrome extension that posed as the AI search engine Perplexity and quietly logged what people searched for. It routed every query and every character typed into the address bar through an attacker-controlled server before redirecting users to real results. Microsoft says Google removed it from the store after responsible disclosure. The extension was called "Search for perplexity ai" (ID flkebkiofojicogddingbdmcmkpbplcd) and used a look-alike domain, perplexity-ai[.]online, to pass for the real service at perplexity.ai. Microsoft's Defender research team  says the point was to intercept searches and collect data. It found no proof of password theft, but far more access than a search box should ever need. Once installed, the extension sets itself as the browser's default search engine. When you searched, the query went first to perplexity-ai[.]online, where the attacker's server logged it with your browser headers, IP address,...

A high-severity flaw in Amazon Q Developer let a malicious repository run commands and steal a developer's cloud credentials. The path was short: a developer opens the repo, trusts the workspace, and Amazon Q does the rest. Amazon has patched it. Tracked as  CVE-2026-12957  (CVSS 8.5), the bug sat in how Amazon's AI coding assistant handled Model Context Protocol (MCP) servers. Wiz Research, which found and reported it, showed that a single config file dropped in a repo was enough to go from git clone to cloud compromise. How the attack worked Amazon Q read an MCP configuration file, .amazonq/mcp.json, from the open workspace and launched the servers it defined. MCP servers are local processes that an AI assistant can spawn to reach databases, APIs, or build tools, so starting one means running commands on the machine. Those processes inherited the developer's full environment. That usually means AWS keys, cloud CLI tokens, API secrets, and SSH agent sockets. ...

AI agents are moving through enterprise environments, inheriting permissions, traversing systems, and executing decisions at machine speed with minimal oversight. The identity infrastructure built to govern human access wasn't designed for autonomous actors, and the gap between what enterprises are deploying and what their governance programs actually cover is widening fast. This guide breaks down how the guardian agents emerged, why it matters, and what operationalizing it looks like in practice. The Governance Gap Agentic AI Created Identity governance has always lagged behind infrastructure change, but the arrival of production-grade agentic AI didn't just widen the gap. It changed its shape entirely. The assumptions baked into every IAM architecture built over the past two decades are no longer sufficient for the environment most enterprises are actually running today. Agents Aren't Service Accounts Security teams have spent years getting reasonably good at go...