artificial intelligence | Breaking Cybersecurity News | The Hacker News

LiteLLM on Friday revealed that an Agent Platform master key accidentally published to a public GitHub repository in late May 2026 ultimately allowed the threat actor to push a malicious version of the popular Python package. "On May 27, 2026, a configuration file was committed to the public LiteLLM-Labs/lite-harness repository which inadvertently contained the master key to a deployment of the LiteLLM Agent Platform, an open source tool we use for orchestrating agents," LiteLLM's Oliver Jensen and Mateo Wang said . "The error was quickly identified, the commit force-pushed over, and we identified the exposed keys as requiring revocation. However, key revocation did not occur for the master key." LiteLLM said the unknown attacker (operating from "69.48.229[.]143") used this unrevoked key on October 10, 2026, to gain access to the Agent Platform and extracted a GitHub CLI OAuth token the service was configured with to allow self-hosted cloud a...

Google has updated its privacy policy to spell out that using personalized artificial intelligence (AI) features in Search or Gemini could allow the company to serve targeted ads to users based on their activity on other connected Google services. "You may choose to allow certain Google services, including services like Search or Gemini Apps, to access data from other Google services you use, such as Google Photos and Workspace apps like Gmail and Drive, in order to provide you with a more personalized experience, subject to the policies and settings of the accessing service," reads a policy update effective October 16, 2026. "For example, Search may process some of the data that you allow Search to access to personalize content, including ads." Google also notes in its revised policy that "We don't show you personalized ads based on your content from Drive, Gmail, or Google Photos, unless you choose to make data from these services available to ...

Cybersecurity researchers have discovered an artificial intelligence (AI)-driven intrusion ecosystem run by affiliates of the INC and The Gentlemen ransomware groups. Central to the operation is an "Autonomous Red Team Kill Chain Framework" referred to as Phantom. The threat actor is assessed to be Russian-speaking and functions both as an intrusion and data-extortion operator, as well as an infrastructure provider supporting different partners across the ransomware ecosystem. "The operators exploited edge and remote-access systems at scale, including FortiClient EMS, SimpleHelp, cPanel, and VPN gateways," SOCRadar said . "They used AI-managed knowledge bases to track victims, stolen credentials, and campaign progress." A notable aspect that defines the campaign is its "AI-executed, human-approved operating structure" in which a human operator provides a victim-specific prompt and reviews every step, while the intrusions themselves are e...

Pillow , the most widely used image library for Python, can be made to run operating system commands on a server by specifying the name of a file it is told to open. The flaw,  CVE-2026-112900 , is in the way Pillow handles EPS images, which it opens by passing the file's name to Ghostscript , a separate program it runs to turn the image into pixels, without checking the name first. Ghostscript can read a name chosen by an attacker as instructions rather than as a file, and point it to a second file that carries commands to run. Pillow's maintainers fixed it in  version 13.0.0 , released on October 16. Pillow turns up almost everywhere Python handles images: web upload forms, avatar and thumbnail tools, document and PDF pipelines, and most machine-learning projects pull it in. OX Security, the firm whose report drew attention to the flaws, estimates the library is downloaded  about 436 million times a month . Version 13.0.0 is a large security release, with fixes...

AI agents paying for API calls with the  x402 SDKs  could sign a payment for their full balance of a supported token if a malicious API asked for it,  security firm V12 said  on October 22. The cause was the SDKs' default client, which set no spending limit until releases on August 18 added one. The risk was limited to agents who automatically signed payments and whose apps set no limits. The missing limit is one of 25 flaws V12 found in the x402 code, and one of three it rated critical and described in detail. The x402 Foundation, which maintains the SDKs, had already fixed all three critical flaws in the August 18 releases. One of the other two let a facilitator, the service that checks x402 payments, accept a Solana transaction that moved no tokens as a valid payment. Among the x402 SDKs, only the TypeScript one had this flaw. The other let a malicious server get an agent to sign a login message for a different website. V12 found the flaws by reviewing t...

Right now, there is probably an AI agent running on a laptop somewhere in your network. It stores API keys and database passwords, can access your internal systems, and works without anyone watching. Nobody reviewed its access. Nobody owns it. If an employee had that much reach with that little oversight, you would call it an insider risk. So why do we still call the agent a productivity tool? Most of the conversation about AI security is about what threat actors are doing with it. But the bigger threat may already be inside your organization.

This week has already been rough for cybersecurity. Critical PHP and GitLab RCE flaws, serious Linux security bugs , compromised eMudhra certificates , a Node.js RCE flaw , a hacked LiteLLM release, and hundreds of hijacked Tor relays . That's a lot to deal with in just a few days. And the week isn't even over yet. But those big headlines are only half the story. Stolen passwords, exposed databases, broken login checks, and trusted tools being turned against their users are causing trouble too. Some attacks are clever. Others work because someone left a simple security gap open. This ThreatsDay, we're digging into the stories that didn't make as much noise but still deserve attention. Some are strange, some are serious, and a few might surprise you. Here's everything else you shouldn't miss this week. The threats change every week. Subscribe, and we’ll alert you when each new ThreatsDay Bulletin is out.

Linux servers running KSMBD, the kernel's built-in SMB server, are exposed to two publicly reported remote code execution flaws rated 9.0 by  Trend Micro's Zero Day Initiative , with no fix yet in the mainline kernel. One of them, ZDI-CAN-33774, can be triggered before any network authentication. The other, ZDI-CAN-33549, requires the SMB Direct transport and an RDMA-capable network interface. ZDI-CAN-33549  is an integer underflow in KSMBD's SMB Direct path. A client can supply RDMA buffer descriptors in which the first is longer than the declared payload, causing the server's size counter to wrap and allowing a subsequent memory copy to overrun the end of an allocated buffer. It requires three things: an RDMA-capable interface (hardware or the rxe software driver), a writable KSMBD share, and a client negotiating SMB Direct. ZDI-CAN-33774  requires no authentication, exposing any server running KSMBD on port 445. The flaw is a race in how KSMBD assign...

Platform-as-a-Service (PaaS) provider Railway on Wednesday disclosed "anomalous activity" related to systems that store customer environment variables. Upon further investigation, the company found that an unauthorized third party may have gained access to encrypted environment variables, and, in some cases, obtained the decrypted contents. The impact, it said, is limited to a "subset of users," although it did not disclose an exact number. Environment variables refer to user-defined key-value pairs that are used to store sensitive data like API keys, database credentials, or feature flags. They are configured at the operating system or container environment level instead of hard-coding the values directly into a program's source code. Railway said the incident is likely part of a series of attacks designed to target platform companies with the ultimate goal of compromising artificial intelligence (AI) inference providers. The company also said it has n...

A flaw in the OpenCode AI coding agent allowed a malicious or hijacked MCP server to run code on a developer's Windows or WSL machine when the developer signed in. OpenCode connects to MCP (Model Context Protocol) servers to add outside tools, and such code would run with the developer's own access. Anomaly , the company behind OpenCode, fixed the flaw in  version 1.18.33  on September 28. It did not request a CVE or publish a security advisory, so vulnerability scanners will not flag the builds still at risk: 1.0.137 through 1.18.32. During that sign-in, the server itself supplies the web address the user is sent to, an OAuth "authorization endpoint." On Windows and WSL, OpenCode handed that address to a helper that opens it in a browser. The helper built a PowerShell command around the address without escaping it, so an address written to carry a PowerShell command would run that command. The browser step runs only when a developer starts the sign-in, by ru...

Microsoft on Wednesday disclosed that a China state-aligned actor dubbed Storm-0940 was behind the recent supply chain compromise of the ConnectWise Cloud Backup application. During the incident, Storm-0940 used ARTEX , a Go-based agentic AI penetration testing framework that was made closed-source in the wake of its abuse in cyber attacks targeting South Korean financial institutions, to speed up scanning and reconnaissance, password spraying, and OAuth application abuse at downstream victims. The tech giant said the threat actor used the AI tool at "machine speed" to automate attacks targeting hundreds of organizations across the world spanning a broad range of sectors since September 2026. Targets include aerospace and defense, automotive, chemicals, energy, financial services, government, legal, manufacturing, media, military, semiconductor, software, technology, and think tanks and non-profits. "Storm-0940's use of ARTEX accelerated the speed of comprom...

The popular LiteLLM Python package has been compromised as part of a software supply chain attack that delivered a credential stealer and a backdoor, according to reports from SafeDep , Socket , StepSecurity , and Upwind . The compromise affects LiteLLM 1.107.0, distributed through PyPI wheels and Docker images. The Python package has since been quarantined on Python Package Index (PyPI), and the malicious Docker images have been removed. Oliver Jensen, director of security at LiteLLM, said an investigation into the incident is ongoing, and that it's still working to trace the original attack vector by which the threat actor published the malicious release. "We have identified that the malicious actions were taken through three engineers' GitHub accounts, appearing to be by way of stolen GitHub CLI OAuth tokens," Jensen said . "The threat actor used one user's account to trigger the project-releaser run, and another user's account to approve it. A...

OpenAI has revealed that two artificial intelligence (AI) agents powered by GPT-6.1 Sol and GPT-6 Astra weakened misalignment monitoring while making requested software changes during two internal coding tasks. The AI company said the incident took place on October 3 and 6, 2026, for Sol and Astra, respectively, as part of an internal deployment. "A GPT-6.1 Sol agent repairing an evaluation pipeline disabled automatic alerts to our Security team, treating an earlier instruction not to send messages during a Slack history search as a reason to suppress monitoring notifications," OpenAI said in a report. The end result of this action was that the misalignment monitor still ran and saved its reports, but automatic human alerts were disabled by setting "no_notify" to true and emptying the lists of people configured to receive the reports. In another task on October 6, a GPT-6 Astra agent migrating a program is said to have removed its monitoring integration in...

Google on Tuesday said it's rolling out improved safeguards in its Chrome browser to secure users against ClickFix attacks when Enhanced Protection is enabled. ClickFix is a popular social engineering technique in which bad actors trick users into copying and pasting malicious code directly into their computer's operating system, thereby bypassing conventional security controls. As part of these attacks, unsuspecting users who land on legitimate-but-compromised sites or click on malicious ads are redirected to phishing landing pages that display fake Cloudflare or Google reCAPTCHA verification or other non-existent errors and instruct users to copy and paste a command on the Windows Run dialog (or the Mac Terminal app) to address it. The moment a user presses the Enter key after pasting the command, it triggers a multi-stage infection chain that leads to the deployment of malware, such as information stealers or remote access trojans. In an attempt to blend in with re...

Generative AI has changed what a malicious email looks like. Messages that once gave themselves away through clumsy wording, generic pretexts, and reused templates now arrive fluent, personalized, and unique to each recipient, and the behavioral signals email filters have relied on for years are fading as a result. Proofpoint shared several examples of the shift at its Protect 2026 conference. In one, a single threat actor contacted more than 30 real estate agents at the same company, each from a different account, about a different property, and in different words, before closing every thread with the same request: join a video meeting. Federal data points the same way. The FBI's Internet Crime Complaint Center counted more than 22,000 complaints involving AI in 2025, with nearly $900 million in associated losses, and logged 24,768 business email compromise (BEC) complaints with about $3.05 billion in reported losses, according to its 2025 Internet Crime Report . What AI C...

An AppSec team can point an AI agent running on a general-purpose model at its own web apps with little setup. Sure, the model writes the attacks. But everything that makes those attacks safe to run against production and worth acting on afterward lives in the harness around the model. The harness is where the engineering in AI pentesting sits. The Hugging Face intrusion OpenAI disclosed in July showed what happens when that harness gives way, even inside an OpenAI evaluation built to contain the agents. Why Each Control Takes Rigorous Engineering Let’s take a closer look at the attack different stages and the role AI agents can take there Scope A pentesting agent makes thousands of decisions about where to send its next request, and the model judges each one against a rule written in plain language. A web app’s real boundary often extends past one domain through redirects and calls to outside services. Enforcing scope means code that inspects every request before it leaves, a...

U.S. federal civilian agencies must remove DeepSeek products from their information technology systems, along with any third-party products or services "developed using DeepSeek implementations," under a Department of Homeland Security order. They may also no longer buy information technology products or services from the Chinese AI developer or its affiliates Homeland Security Secretary Markwayne Mullin issued the order, dated July 24, 2026, under the  Federal Acquisition Supply Chain Security Act of 2018 , and the department published it in the  Federal Register  on October 20. The notice does not define " DeepSeek implementations. " The order names Hangzhou DeepSeek Artificial Intelligence Co., Ltd., the developer of the R1 and V3 models, and covers its affiliates as defined in federal acquisition rules. The company's  privacy policy  says it collects, processes and stores users' personal data in the People's Republic of China. Under the 2018...

A Chinese-speaking criminal pointed autonomous AI agents at company websites and let them break in with almost no help, and a security researcher watched it happen because the operators left their control panels open on the internet. The agents tried more than 1,200 sites and reached websites belonging to more than 100 companies, including farm-equipment maker Deere, appliance company SharkNinja and engine maker Cummins,  according to Reuters , which reviewed the attacker's own records to identify them. How much data, if any, was taken is not known. The researcher, Snorre Fagerland of the security firm NorForge , found the operations by scanning the internet for exposed copies of  Cairn , an open-source AI tool popular with hackers. Beginning around October 2, he found about two dozen Cairn instances being used maliciously, which let him read the attackers' work in real time. He set out the findings on Monday, and Reuters, which saw the report before publication, named...

WhatsApp's help pages carried hidden instructions for its support chatbot that told it not to recommend a data protection authority or regulator to users who could no longer delete their accounts, even though WhatsApp's own privacy policy for Europe tells users they can complain to one. The text sat in the pages' code, never shown on screen, until it was removed by October 18. Whether the chatbot ever followed the instructions is not known:  Redact , the data-deletion software company that found the text, read the instructions, not the chatbot's answers. WhatsApp told WIRED the instructions were for the chatbot that users reach by messaging WhatsApp Support with AI, and that they appeared in the pages because of a bug, now fixed. A second block of hidden text on the page about reporting and blocking told the chatbot that once a user mentioned a lawyer, a lawsuit, or a court, it should not cite privacy laws such as the GDPR, even when the user raised them. A W...

A sales manager builds an AI-assisted workflow to route leads. It connects to customer records, summarizes inbound requests, and posts updates to a shared channel. The workflow saves time, reduces manual work, and soon becomes part of the team’s daily routine. Across the same organization, employees are building useful apps, agents, and automations faster than existing processes can support them.  If you work in security, that success introduces a new set of questions. Whose credentials does the workflow use? Which records can it access? Who can see its output? If it starts exposing information it shouldn’t, can you trace what happened and stop further access?  Wild code takes root when useful builds spread faster than the processes for owning, securing, and maintaining them. It isn’t simply another name for unapproved software. Employees are creating new capabilities, not just adopting existing products. A small experiment can become an operational dependency before secu...