Web Security | Breaking Cybersecurity News | The Hacker News

Three researchers at the security firm  Hacktron  used Anthropic's Claude Opus 5 to chain two flaws and take over the ChatGPT and Codex accounts of several OpenAI employees, then reach an internal OpenAI code repository. The chain began with a bug in the software that runs OpenAI's public help forum and moved through a weakness in OpenAI's own login system. This was security research, not a real-world attack: the team reported the flaws to OpenAI, proved the access with a harmless pull request, and then stopped. From the first look, that internal access took under 72 hours. OpenAI confirmed a fix about 14 hours after the report, according to Hacktron, and on September 1 paid the team a $6,500 bounty. OpenAI said the award "recognizes the OpenAI-side finding, not the actions against Discourse," the open-source software that runs the forum. Testing the forum itself was outside its bug bounty program. OpenAI has not publicly described the login flaw, and it ...

A critical vulnerability impacting Orkes Conductor is being actively exploited in the wild, according to Fortinet. The vulnerability in question is CVE-2026-58138 (CVSS v3.1 score: 9.8/CVSS v4 score: 9.3), which relates to a case of unauthenticated remote code execution. "Orkes Conductor 3.21.21 before 3.30.2 contains an unauthenticated remote code execution vulnerability that allows remote attackers to execute arbitrary OS commands by submitting inline workflow definitions containing malicious JavaScript or Python expressions to the workflow API endpoint prior to authentication," a description of the flaw on the NIST National Vulnerability Database (NVD) reads. "Attackers can exploit unsandboxed GraalVM evaluators configured with HostAccess.ALL or allowAllAccess(true) through INLINE, LAMBDA, DO_WHILE, and SWITCH task types to invoke arbitrary system commands via Java reflection or direct subprocess calls." In an outbreak alert issued this week, Fortinet ...

Google's Gemini model has become the latest artificial intelligence (AI) system to access the internet and break into other companies during a cybersecurity evaluation. The development was first reported by The Wall Street Journal. The incidents occurred in May 2026 as part of a test run conducted by Israeli company Irregular. The evaluation partner was also involved in similar hacks disclosed by OpenAI, Anthropic, and Meta. According to the Journal, the model gained access to a protected system after repeatedly guessing its password. Two other cases related to the model finding credentials in a public repository, allowing it to obtain unauthorized access to protected systems. However, unlike other incidents observed in the case of Anthropic and OpenAI, the Gemini model ended the intrusion after finding that it had breached a real company's system. Irregular is said to have notified Google of the incidents in July 2026. In a report published last month, Irregular pinn...

WordPress today released patches to fix a new set of vulnerabilities in its core software, one of which could allow a crafted web link, opened by a logged-in administrator, to install a theme from the official WordPress.org directory without anyone clicking Install. The security firm pwn.ai , whose researchers reported the flaw, calls the attack chain Click2Shell . On its own the flaw only installs a real theme that the attacker picks, but the security research team showed it could be combined with a separate weakness in a theme to run the attacker's own code on the server.  The fix shipped on September 17 in WordPress 7.1.1 . Because this is a security release, WordPress advises updating right away. There is no sign the flaw has been used in real attacks. The installed theme stays switched off, so the site's own appearance does not change and nothing looks wrong. Reaching code execution needed a second, separate flaw in the theme that was installed. As pwn.ai wrote of t...

In July 2025, someone registered a domain that used to belong to a content delivery network.  The CDN had been wound down years earlier, and the domain it served assets from was allowed to expire. What it had not lost were its callers. Thousands of websites, code repositories, and documentation pages still carry hard-coded references to hostnames beneath it.  The new owner holds wildcard DNS across the entire domain, and any hostname under it now resolves to infrastructure that person controls. Today the apex serves an ad-heavy media downloader page, which is unremarkable. The remarkable part is that the decision about what those thousands of pages load next belongs to a stranger, and nobody involved has been notified, because from the outside nothing broke.  This pattern is not hypothetical, and it's not unheard of either. In June 2024, the polyfill.io domain (a JavaScript shim embedded in more than 110,000 sites) changed ownership and began serving conditional redi...

Cybersecurity researchers have discovered a cluster of 13 npm packages that have been found to deliver a previously undocumented JavaScript stealer codenamed WeaselBiscuit . The new malware family, per OpenSourceMalware , exhibits functional overlaps with two malware strains associated with the Democratic People's Republic of Korea's (DPRK) Contagious Interview campaign: BeaverTail and OtterCookie . "It's smaller, lighter, and stripped down, with many of the heavier functions removed entirely," security researcher Paul McCarty (aka 6mile) said. The names of the packages are below - @biz44/id10-client @biz44/id12-client @biz44/id44-client @biz44/id79-client @biz44/id95-client @biz44/id99-client @biz44/process-runtime-utils @biz44/runtime-utils engin1 id79-client process-lhpm process-mite process-tailwind "It's a stripped down stealer that borrows several functions from DPRK's BeaverTail and OtterCookie,...

Attackers now weaponize new vulnerabilities in about five days (Mandiant, part of Google Cloud). The median organization takes 43 days to patch one (Verizon DBIR 2026). A new free guide explains how autonomous AI agents are closing that gap, and what security leaders must demand before pointing one at production. TL;DR Exploitation is now the front door. It starts 31% of breaches (Verizon DBIR 2026), the #1 initial-access vector, while annual pentesting leaves an estimated 90% of the estate untested. The capability is proven, not projected. An autonomous system topped HackerOne's US leaderboard in 2025 (XBOW), and peer-reviewed agents exploited 87% of one-day flaws unaided (Fang et al., 2024). Continuous beats periodic, measurably. Programmatic testing makes teams 4.5x more likely to fix criticals within three days (Cobalt, 2026). It's an AI agent in your production. Hold it to a bar. Provable coverage, an independent validator, blast-radius guardrails, and a...

A security breach at Gyazo , Helpfeel's image-sharing service, exposed about 23.62 million user records, including email addresses and password hashes, the Kyoto-based company said in a  notice  published Wednesday. It also exposed about 490 million image metadata records, mostly for images from January 2019 or earlier, including the IDs that make up Gyazo image links. Helpfeel said those IDs could be used to view the images without permission, and that it has temporarily disabled viewing of some of them. Helpfeel asked every Gyazo user to change their password and to change it on any other service that uses the same or a similar one. It also asked users to watch for suspicious emails or messages related to the incident. The attacker gained access through a vulnerability in Gyazo's image upload server, ran arbitrary commands on Helpfeel's systems, and accessed Gyazo's database, the company said. It has not said what kind of flaw it was. Helpfeel said no pay...

Cisco has warned of a fresh maximum-severity security flaw impacting Identity Services Engine (ISE) that has come under active exploitation. The vulnerability, tracked as CVE-2026-76460 (CVSS score: 10.0), could allow an unauthenticated, remote attacker to bypass authentication. "This vulnerability is due to insufficient authentication control on an API endpoint," Cisco said. "An attacker could exploit this vulnerability by sending a crafted request to an affected API endpoint. A successful exploit could allow the attacker to gain unauthorized access to the affected device by bypassing the web-based management interface." The issue affects Cisco ISE and Cisco ISE Passive Identity Connector (ISE-PIC), regardless of device configuration. It has been addressed in the following versions -

A critical security flaw in Issabel Framework , a web-based framework for the open-source unified communications PBX software, has come under active exploitation. The vulnerability in question is CVE-2026-89026 (CVSS v3.1 score: 9.8/CVSS v4.0 score: 9.3), which can allow an unauthenticated remote attacker to execute arbitrary operating system (OS) commands by taking advantage of a hard-coded JSON Web Token (JWT) signing key. The Issabel Framework "contains a hard-coded HS256 JWT signing key in the pbxapi index.php file that is identical across every installation, allowing unauthenticated remote attackers to forge valid bearer tokens," VulnCheck said in an alert. "Attackers can use the forged token to call the manager '/pbxapi/manager/originate' endpoint with the System application parameter, causing Asterisk to execute arbitrary OS commands as the Asterisk user." A patch for the vulnerability was pushed on August 1, 2026, and plugs the flaw by r...

Security researchers at Forever Security have shown that one ordinary browser extension could take control of the AI assistants built into five Chromium-based products: Gemini Live in Chrome, Perplexity Comet, Microsoft Edge, Opera Neon and the Claude in Chrome extension. Once the extension was installed, it could access each product's built-in AI with a single click. On Comet, Edge, Opera Neon, and Claude in Chrome, it could drive the AI agent to act on behalf of the attacker; on Chrome and Comet, it could read files from the user's computer, and on Chrome, it could also switch on the camera and microphone. The findings are researcher demonstrations, not attacks seen in the wild, and each requires the attacker's extension to be already running in the victim's browser. These products all work the same way. The AI has a "body" inside the browser that can see the screen, open files, use the camera, and take actions, and a "brain" that runs on th...

N0va is targeting organizations across North America and Europe with phishing campaigns that impersonate trusted services and abuse legitimate authentication flows. Successful attacks can give threat actors access to valid accounts without relying on obvious malware activity. From there, a single compromised identity can open the door to sensitive data, business systems, and additional cloud resources. The longer that access goes unnoticed, the greater the potential for wider compromise, operational disruption, and financial loss. N0va Is Reaching Organizations Across High-Risk Sectors N0va activity has been observed across organizations in government, technology, consulting, healthcare, and other sectors in North America and Europe. Its use of trusted business platforms and cloud services makes the campaign relevant across a wide range of organizations. N0va phishing campaign attack details Related activity can be traced in ANY.RUN’s Threat Intelligence Lookup using a chara...

Threat actors are exploiting a critical security flaw in WooCommerce Wholesale Lead Capture, a premium WordPress plugin that has more than 6,000 active installs. "This vulnerability can be leveraged by unauthenticated attackers to upload arbitrary files, including PHP backdoors, and achieve remote code execution," Wordfence said . The WordPress security company said it has blocked over 100,000 exploit attempts targeting the vulnerability since June 2026, with 99 of those attack attempts recorded over the past 24 hours. The vulnerability is tracked as CVE-2026-27540 (CVSS score: 9.8). The issue is a case of arbitrary file uploads due to missing file type validation in an AJAX action named "wwlc_file_upload_handler" that impacts all versions of the plugin up to, and including, 2.0.3.1. This opens the door for an unauthenticated attacker to upload arbitrary files on the affected site's server, paving the way for remote code execution. In the attack obse...

A critical security flaw in WSO2 API Manager has come under active exploitation in the wild, according to findings from watchTowr. The vulnerability, tracked as CVE-2026-5430 (CVSS score: 9.8/10.0), is a case of improper verification of a cryptographic signature that could result in account takeover. Hacktron Team has been credited with discovering and reporting the flaw. "JWT authentication can be bypassed when a token is signed using an unsupported algorithm, allowing unauthorized access," according to an advisory released by WSO2 in May 2026. "Successful exploitation of the vulnerability may lead to unauthorized access, including potential compromise of administrative accounts and full account takeover." The shortcoming affects the following products - WSO2 API Control Plane: 4.6.0, 4.5.0 WSO2 API Manager: 4.6.0, 4.5.0, 4.4.0, 4.3.0, 4.2.0, 4.1.0 WSO2 Traffic Manager: 4.6.0, 4.5.0 WSO2 Universal Gateway: 4.6.0, 4.5.0 Fixes are available...

Cybersecurity researchers have shed light on a previously undocumented Brazilian banking malware operation that delivers a toolkit called KREMLIN . Elastic Security Labs is tracking the activity under the moniker REF9334 . Active since at least May 2025, the threat actor has used lures that impersonate a dozen Brazilian banks and install a malicious browser extension on Google Chrome and Microsoft Edge. "The KREMLIN malware ecosystem employs multi-stage JavaScript loaders, custom C++ installers, and malicious browser extensions to steal credentials, session tokens, and sensitive data," security researchers Cyril François and Andrew Pease said in a technical report shared with The Hacker News. "Malicious browser extensions bypass Chromium integrity mechanisms by manipulating Secure Preferences and regenerating required HMACs, and App-Bound encrypted hashes." A defining aspect of the operation is the use of blockchain to conceal the threat actor-controlled...

With artificial intelligence (AI) shrinking the window between vulnerability discovery and exploitation and lowering the barrier to entry for bad actors, new findings from Sysdig show that skilled human operators can move just as swiftly after gaining initial access. In one instance highlighted by the cloud security company, the threat actor pivoted from a vulnerable Marimo notebook to an SSH bastion host in eight seconds using a custom Python toolkit they "wrote and debugged by hand" without any AI agent in the loop. "Eight seconds is the kind of speed we expect to see in AI-assisted attacks," the Sysdig Threat Research Team said . "This operator got there on skill alone, and along the way walked straight past a trap that every agentic threat actor (ATA) we've profiled against this same CVE fell into. Not only can skilled human attackers move at machine speed, but they can also often better evade defenders' detections." The attack chain ha...

Cybersecurity researchers have disclosed details of a mass-scanning campaign that has targeted Vite deployments siphon sensitive data. The first is an automated effort aimed at internet-exposed Vite development servers that's designed to steal cloud credentials, configurations from Amazon Web Services (AWS) and Microsoft Azure instances, and infrastructure state files, per F5 Labs . The credential harvesting activity, observed in August 2026, has been found to leverage an exploit for CVE-2026-39364 (CVSS score: 8.2), a high-severity security flaw in Vite that could permit an unauthenticated attacker to bypass security restrictions via query parameter manipulation and leak sensitive data, including files specified by server.fs.deny. "On the Vite dev server, files that should be blocked by server.fs.deny (e.g., .env, *.crt) can be retrieved with HTTP 200 responses when query parameters such as ?raw, ?import&raw, or ?import&url&inline are appended," Vite s...

A critical vulnerability in LiteSpeed Web Server Enterprise could let a low-privilege website user gain root access on a shared-hosting server, cPanel warned in an  advisory published on September 14 . On such servers, many customers' sites run on a single machine, and an attacker with one of those hosting accounts could exploit the flaw to access or alter other sites and the server itself, according to the advisory. cPanel said it had received notice of the flaw, which affects versions before 6.3.7, and urged administrators to update to that release, which LiteSpeed  published on September 11 . The flaw can bypass the controls that keep hosting accounts apart, including  CageFS , cPanel said. CageFS is a CloudLinux tool that gives each hosting account a restricted view of the file system, so it cannot see other accounts or the server's configuration files. Neither cPanel's advisory nor LiteSpeed's release notes describe how the flaw works. LiteSpeed's an...

A Chinese threat actor has been attributed to a spear-phishing campaign that exploits recently patched security flaws in Google Chrome and Microsoft Windows to deliver a malicious JavaScript backdoor called GRIMWEDGE . Volexity, which is tracking the threat cluster under the moniker UTA0560 , said the activity targeted multiple non-governmental organizations (NGOs) on September 1, 2026. "The emails contained a message encouraging the users to click a link that led to the website of a U.S.-based university," researchers Ankur Saini, Conor Quigley, Sean Koessel, Steven Adair, and Tom Lancaster said . "These links abused a reflected cross-site scripting (XSS) vulnerability on the website, redirecting recipients to threat-actor-controlled infrastructure hosting a multi-stage exploit chain." The exploit chain, as previously highlighted by Proofpoint, involves three separate flaws – two in Chrome and one in Windows Advanced Local Procedure Call (ALPC). It first ab...

A flaw in Telegram Desktop let a bot's message plant hidden JavaScript inside chats that users exported to HTML files, security researchers at ExPatch said in a  writeup  published on September 12. In Telegram, the message looked ordinary, with a link button, and the script ran only when someone opened the export file in a web browser. It could then copy every message in that file to an attacker-controlled server, or rewrite what the page displayed. Telegram shipped  a fix  in July, but the app update does not update files exported with earlier versions, so old HTML exports can still carry the script. Telegram Desktop, Telegram's app for Windows, macOS, and Linux, can save a single chat or all chats from an account as HTML pages that open in a browser. Bots can attach rows of buttons under their messages, which Telegram calls inline keyboards, and the bot chooses the text shown on each button. Until the fix, the export code wrote that button text directly ...