Application Security | Breaking Cybersecurity News | The Hacker News

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...

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...

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...

Threat actors have been observed attempting to exploit a recently patched critical security flaw in Gitea Docker images, according to Sysdig . The vulnerability in question is CVE-2026-20896 (CVSS score: 9.8), a vulnerability that stems from the DevOps platform trusting the "X-WEBAUTH-USER" header from any source IP address, effectively allowing an unauthenticated internet client to get elevated access. In a statement shared with The Hacker News via email, security researcher Ali Mustafa (@rz1027), who is credited with discovering and reporting the flaw, said the Gitea Docker images shipped an "app.ini" template that hard-codes "REVERSE_PROXY_TRUSTED_PROXIES = *" by default. The " app.ini " file is a core configuration file for managing server parameters, database connections, security behavior, and application settings. "With reverse-proxy login enabled, that wildcard trusts every source IP, so anyone who could reach the port could...

Researchers tested 444 AI chatbot apps for iPhone and found that 282 of them, nearly two-thirds, exposed paid AI access through their network traffic. In many cases, the path in was visible just by watching what the app sent: a plaintext API key, a reusable token, or a backend server that accepted requests with no key at all. Whoever grabs it can send model requests on the developer's account, and the developer pays the bill. Three months after the researchers warned the developers, only 28% had fixed it. The work, from researchers at Wake Forest University, is the  first in-depth study of the problem on iOS . It is striking partly because of how little effort the snooping took. The team used a tool they built, LLMKeyLens , that watches an app's traffic and pulls out the credentials as they go by. No jailbreaking, no cracking the app open. The key is the secret that lets the app call a service like OpenAI or Google Gemini. Embed it in the app, and it is exposed with ev...

Cybersecurity researchers have described what they say is a new class of attack that can trick artificial intelligence (AI) coding agents into running arbitrary code on developer machines. Called Agentjacking by Tenet Security, the attack can be triggered by means of a fake error report crafted using Sentry, an open-source error-tracking and performance-monitoring platform. "The attack exploits a critical architectural flaw at the intersection of Sentry's event ingestion (which accepts arbitrary payloads from anyone with the DSN) and the Sentry MCP server (which returns this data to AI agents as trusted system output)," security researchers Ron Bobrov, Barak Sternberg, and Nevo Poran said . The idea is to inject crafted input into Sentry error events, which are then interpreted by coding agents like Claude Code and Cursor as legitimate diagnostic resolution steps and run attacker-controlled code. A successful attack of this kind can expose sensitive data, includ...

Redis has patched a use-after-free in its blocking-client code that lets an authenticated user run arbitrary OS commands on the machine hosting the database. The flaw was found by an autonomous AI tool built to hunt bugs in large codebases. Tracked as CVE-2026-23479 , the flaw was introduced in Redis 7.2.0 and remained in every stable branch until the May 5 fixes, unnoticed for over two years. NVD rates it 8.8 under CVSS 3.1; Redis lists it as 7.7 under CVSS 4.0. It was reported by Team Xint Code, and a complete technical  write-up is now public. The cloud footprint makes this worse. Wiz's analysis, published with the exploit writeup, puts Redis in a large majority of cloud environments, with most of those instances running without a password. The exploit needs an authenticated session, but in a default deployment, the default user already holds every privilege the chain requires. The flaw lives in unblockClientOnKey() in src/blocked.c , which fires when a key event wakes ...

Shadow AI used to mean employees pasting things they shouldn't into ChatGPT. It now means something bigger: employees building full applications with AI, wiring them into production systems, and publishing them on the open internet. Without Security or IT in the loop. The artifact moved from a prompt to a product. The risk surface moved with it. In The Shadow Builders report ( get it here ), a new category-level investigation covered in May by Axios, WIRED, and VentureBeat, Red Access identified more than 380,000 publicly accessible web assets across the leading vibe-coding platforms. Roughly 5,000 looked corporate. More than 2,000 of those held sensitive corporate, operational, or personal data - sitting on the open web, deployed without basic access controls, often granting admin access by default to anyone who reached the URL. Six continents. Every industry is examined. No exploitation required. Inside organizations, passing their audits while these exposures were live...

Cybersecurity researchers have disclosed multiple security vulnerabilities impacting NGINX Plus and NGINX Open, including a critical flaw that remained undetected for 18 years. The vulnerability, discovered by depthfirst , is a heap buffer overflow issue impacting ngx_http_rewrite_module (CVE-2026-42945, CVSS v4 score: 9.2) that could allow an attacker to achieve remote code execution or cause a denial-of-service (DoS) with crafted requests. It has been codenamed NGINX Rift . "NGINX Plus and NGINX Open Source have a vulnerability in the ngx_http_rewrite_module module," F5 said in an advisory released Wednesday. "This vulnerability exists when the rewrite directive is followed by a rewrite, if, or set directive and an unnamed Perl-Compatible Regular Expression (PCRE) capture (for example, $1, $2) with a replacement string that includes a question mark (?)." "An unauthenticated attacker, along with conditions beyond its control, can exploit this vulnerabili...

TL;DR: Stop chasing thousands of "toast" alerts. Join experts from Wiz to learn how hackers connect tiny flaws to build a "Lethal Chain" to your data—and how to break it. Register for the Strategic Briefing Here . Most security tools work like a smoke alarm that goes off every time you burn a piece of toast. You get so many alerts that you eventually start to ignore them. The real danger? While your team is busy fixing 100 "toast" alerts, a sophisticated attacker is quietly building a Lethal Chain through your system. Hackers rarely look for one big "open door" anymore. Instead, they find a series of tiny, low-risk "cracks" that don't look scary on their own. By connecting these cracks—moving from a small coding bug to a cloud misconfiguration—they create a direct path to your most sensitive data. If your tools only look at code or cloud in isolation, you aren’t seeing the big picture. You’re flying blind. The Briefing: Sto...

Google has announced expanded Binary Transparency for Android as a way to safeguard the ecosystem from supply chain attacks. "This new public ledger ensures the Google apps on your device are exactly what we intended to build and distribute," Google's product and security teams said . The initiative builds upon the foundation of Pixel Binary Transparency , which Google introduced in October 2021 to bolster software integrity by ensuring that Pixel devices are only running verified operating system (OS) software by keeping a public, cryptographic log that records metadata about official factory images. The verifiable security infrastructure mirrors Certificate Transparency , an open framework that requires all issued SSL/TLS certificates to be recorded in public, append-only, and cryptographically verifiable logs to help detect mis-issued or malicious certificates. The move is aimed at countering the risks posed by binary supply chain attacks, which often deliver ...

Bitwarden CLI , the command-line interface for the password manager Bitwarden, has reportedly been compromised as part of a newly discovered and ongoing Checkmarx supply chain campaign , according to findings from JFrog and Socket. "The affected package version appears to be @bitwarden/cli@2026.4.0 , and the malicious code was published in 'bw1.js,' a file included in the package contents," the application security company said . "The attack appears to have leveraged a compromised GitHub Action in Bitwarden's CI/CD pipeline, consistent with the pattern seen across other affected repositories in this campaign." In a post on X, JFrog said the rogue version of the package "steals GitHub/npm tokens, .ssh, .env, shell history, GitHub Actions and cloud secrets, then exfiltrates the data to private domains and as GitHub commits." Specifically, the malicious code is executed by means of a preinstall hook, resulting in the theft of local, CI, Git...

Imagine a world where hackers don't sleep, don't take breaks, and find weak spots in your systems instantly. Well, that world is already here. Thanks to AI, attackers are now launching automated, large-scale exploits faster than ever before. The time you have to fix a vulnerability before it gets attacked is shrinking to zero. We call this the Collapsing Exploit Window , and it means your standard patching routine is officially too slow. If you are fighting AI-speed attacks with manual-speed defenses, your systems are at a breaking point. It’s time to rethink everything. Join our highly anticipated webinar featuring expert guest Ofer Gayer, Vice President of Product at Miggo Security, and learn how to beat the bots at their own game: Mythos and the Collapsing Exploit Window: Rethink Vulnerability Prioritization at AI Speed . Here is exactly what you will walk away with: The Truth About Mythos: We are cutting through the hype. Learn what Mythos actually represents and w...

Cybersecurity researchers have discovered a critical "by design" weakness in the Model Context Protocol's ( MCP ) architecture that could pave the way for remote code execution and have a cascading effect on the artificial intelligence (AI) supply chain. "This flaw enables Arbitrary Command Execution (RCE) on any system running a vulnerable MCP implementation, granting attackers direct access to sensitive user data, internal databases, API keys, and chat histories," OX Security researchers Moshe Siman Tov Bustan, Mustafa Naamnih, Nir Zadok, and Roni Bar said in an analysis published last week. The cybersecurity company said the systemic vulnerability is baked into Anthropic's official MCP software development kit (SDK) across any supported language, including Python, TypeScript, Java, and Rust. In all, it affects more than 7,000 publicly accessible servers and software packages totaling more than 150 million downloads. At issue are unsafe defaults in...

A "novel" social engineering campaign has been observed abusing Obsidian, a cross-platform note-taking application, as an initial access vector to distribute a previously undocumented Windows remote access trojan called PHANTOMPULSE in attacks targeting individuals in the financial and cryptocurrency sectors. Dubbed REF6598 by Elastic Security Labs, the activity has been found to leverage elaborate social engineering tactics through LinkedIn and Telegram to breach both Windows and macOS systems, approaching prospective individuals on the professional social network under the guise of a venture capital firm and then moving the conversation to a Telegram group where several purported partners are present. The Telegram group chat is engineered to lend the operation a smidgen of credibility, with the members discussing topics related to financial services and cryptocurrency liquidity solutions. The target is then instructed to use Obsidian to access what appears to be a s...

OpenAI on Tuesday unveiled GPT-5.4-Cyber , a variant of its latest flagship model, GPT‑5.4 , that's specifically optimized for defensive cybersecurity use cases, days after rival Anthropic unveiled its own frontier model, Mythos . "The progressive use of AI accelerates defenders – those responsible for keeping systems, data, and users safe – enabling them to find and fix problems faster in the digital infrastructure everyone relies on," OpenAI said . In conjunction with the announcement, the artificial intelligence (AI) company said it's ramping up its Trusted Access for Cyber ( TAC ) program to thousands of authenticated individual defenders and hundreds of teams responsible for securing critical software. AI systems are inherently dual-use, as bad actors can repurpose technologies developed for legitimate applications to their own advantage and achieve malicious goals. One core area of concern is that adversaries could invert the models fine...

OX Security recently analyzed 216 million security findings across 250 organizations over a 90-day period. The primary takeaway: while raw alert volume grew by 52% year-over-year, prioritized critical risk grew by nearly 400%. The surge in AI-assisted development is creating a "velocity gap" where the density of high-impact vulnerabilities is scaling faster than remediation workflows. The ratio of critical findings to raw alerts nearly tripled, moving from 0.035% to 0.092%. Key Findings from the 2026 Analysis: CVSS vs. Business Context: Technical severity scores are no longer the primary driver of risk. The most common elevation factors were High Business Priority (27.76%) and PII Processing (22.08%) . In modern environments, where a vulnerability lives is now more important than what the vulnerability is. The AI Fingerprint: We observed a direct correlation between the adoption of AI coding tools and the quadrupling of critical f...

Adobe has released emergency updates to fix a critical security flaw in Acrobat Reader that has come under active exploitation in the wild. The vulnerability, assigned the CVE identifier CVE-2026-34621 , carries a CVSS score of 8.6 out of 10.0. Successful exploitation of the flaw could allow an attacker to run malicious code on affected installations. It has been described as a case of prototype pollution that could result in arbitrary code execution. Prototype pollution refers to a JavaScript security vulnerability  that permits an attacker to manipulate an application's objects and properties. The issue impacts the following products and versions for both Windows and macOS - Acrobat DC versions 26.001.21367 and earlier (Fixed in 26.001.21411) Acrobat Reader DC versions 26.001.21367 and earlier (Fixed in 26.001.21411) Acrobat 2024 versions 24.001.30356 and earlier (Fixed in 24.001.30362 for Windows and...

The Fragmented State of Modern Enterprise Identity Enterprise IAM is approaching a breaking point. As organizations scale, identity becomes increasingly fragmented across thousands of applications, decentralized teams, machine identities, and autonomous systems.  The result is Identity Dark Matter: identity activity that sits outside the visibility of centralized IAM and beyond the reach of security teams. According to Orchid Security ’s analysis , 46% of enterprise identity activity occurs outside centralized IAM visibility. In other words, nearly half of the enterprise identity surface may be operating unseen. This hidden layer includes unmanaged applications, local accounts, opaque authentication flows, and over-permissioned non-human identities. It is further amplified by disconnected tools, siloed ownership, and the rapid rise of Agentic AI. The consequence is a widening gap between what the security organizations think th...

Cybersecurity researchers have disclosed details of a now-patched bug impacting Open VSX's pre-publish scanning pipeline to cause the tool to allow a malicious Microsoft Visual Studio Code (VS Code) extension to pass the vetting process and go live in the registry. "The pipeline had a single boolean return value that meant both 'no scanners are configured' and 'all scanners failed to run,'" Koi Security researcher Oran Simhony said in a report shared with The Hacker News. "The caller couldn't tell the difference. So when scanners failed under load, Open VSX treated it as 'nothing to scan for' and waved the extension right through." Early last month, the Eclipse Foundation, which maintains Open VSX, announced plans to enforce pre-publish security checks before VS Code extensions are published to the repository in an attempt to tackle the growing problem of malicious extensions. With Open VSX also serving as the extension market...