artificial intelligence | Breaking Cybersecurity News | The Hacker News

A blank field. A public repo. One reply to an email. A box left exposed. None of this sounds dramatic, which is partly the problem. This week’s threats keep finding leverage in small things that were easy to overlook. There are actively exploited bugs in the mix, cleaner intrusion paths, smarter automation, and a long patch list waiting behind them. Some attacks are getting more capable. Others are still getting in because the basics gave way first. Here’s what mattered this week. ⚡ Threat of the Week Citrix Warns of Newly Exploited NetScaler ADC and Gateway Flaw — Citrix released security updates for a high-severity security flaw in NetScaler ADC and NetScaler Gateway that has been exploited as part of targeted zero-day attacks. The vulnerability, tracked as CVE-2026-88779, carries a CVSS score of 8.7 out of 10.0. "CVE-2026-88779 is a memory overflow vulnerability in Citrix NetScaler ADC and Citrix NetScaler Gateway that can lead to denial-of-service under specific depl...

Every modern enterprise depends on credentials. This is how humans, systems, and now AI, all connect to data, services, and each other securely. GitGuardian helps secure that credential layer through three connected capabilities: Detect, Remediate, and Prevent. The journey starts with detection, because organizations first need to understand what credentials exist, where they live, and what they can access. This is the first of three articles we are releasing that explain the reason behind our mission.  — Software production is accelerating beyond the growth assumptions that shaped many of today's security controls. GitHub COO Kyle Daigle said the platform had gone from roughly 1 billion commits during all of 2025 to 2.9 billion commits in August 2026, an annualized pace of over 14 billion for the 2026 reporting year. GitHub's own engineering team has gone further in its capacity planning, saying it moved from preparing for 10x scale to designing for a future that require...

Apple has announced that it's taking steps to tighten controls around a macOS setting called Full Disk Access (FDA) due to security risks posed by artificial intelligence (AI) agents. "Some developers are using Full Disk Access in ways that could put users at risk, exposing everything on their systems—including files, mail, messages, and even browsing history – without users' full knowledge and understanding," Apple said in a post. "For communication apps, this can also compromise the privacy of the people users are communicating with." Full Disk Access , accessed via Privacy & Security in the Settings app, was introduced by Apple in macOS Mojave (version 10.14), offers users greater control over which applications can access their entire system and data from apps like Mail, Messages, Safari, and Time Machine backups. Once the setting is enabled for an application, it allows that program to bypass certain security restrictions and read and writ...

A critical security flaw impacting Rejetto HTTP File Server (HFS) is witnessing active exploitation attempts, according to VulnCheck. The vulnerability in question is CVE-2026-61500 (CVSS score: 9.3), a case of session forgery stemming from the use of a weak pseudo-random number generator (PRNG) that can lead to a predictable key, which an attacker can then use to gain unauthorized access and seize control of affected systems. "Rejetto HFS 3.0.0 through 3.2.0 derives its session-cookie signing key from the non-cryptographic Math.random() generator and discloses outputs of the same generator to unauthenticated clients during login," according to an advisory for the flaw. "A remote attacker can collect a small number of login responses, reconstruct the generator's state, recover the signing key, and forge a valid administrator session cookie, leading to full administrative access and remote code execution via the server_code configuration feature." Hori...

A new China-nexus cyber espionage group known as TA419 has been attributed to multiple credential phishing campaigns targeting artificial intelligence (AI) experts working for U.S. think tanks, universities, and legal sector organizations. The campaigns have impersonated prominent economists and AI policymakers, as well as a prominent Anthropic employee, to single out an AI policy expert at a U.S. think tank in February 2026. The phishing email carried the subject line "Request for Feedback on Military Integration of Claude." "This activity likely supports wider Chinese intelligence objectives to better understand ongoing developments within the U.S. AI policy and regulatory landscape and occurs amid intense strategic competition, accusations of model distillation, and export controls involving the U.S. and China," Proofpoint said in an analysis published this week. The enterprise security company has described TA419 as a China-aligned and espionage-motivat...

The U.K.'s domestic intelligence and security agency has warned that more than 100 academics have helped China boost its intelligence gathering efforts on behalf of Beijing's state security service. In a "Security Service Espionage Alert" issued on September 30, 2026, MI5 said the "primary purpose of the China General Technology Research Institute (CGTRI) 中国通用技术研究院 is to fund research that directly improves Chinese Ministry of State Security (MSS) technical capability for espionage." CGTRI, also known as the China Academy of General Technology (CAGT), is assessed to be a front company for MSS. The body, per MI5, funds academic research in China on topics including artificial intelligence (AI), cybersecurity, covert communications systems, and steganography. More than 100 U.K.-linked academics have contributed to research projects funded by MSS via CGTRI, the alert read. In some cases, the individuals may not be aware that CGTRI is providing monetary ...

Featuring: Cybersecurity is being reshaped by the expansion of cloud infrastructure, AI, distributed systems, and increasingly complex digital environments. As organizations manage more identities, devices, data, and internet-facing infrastructure, security is shifting toward continuous visibility, control, and the ability to respond to risk at scale. This report examines how core areas of cybersecurity are evolving in response to that shift. Across identity security, telemetry management, human security, endpoint management, human risk intelligence, exposure management, email and domain security, connected device security, AI-native security operations, and cloud security, it explores how organizations are adapting to threats that increasingly move across systems, identities, and infrastructure rather than targeting a single point of failure. Read the full report here:  https://report.papryon.com/thehackernews

A critical flaw in GitLab's AI Gateway could let a logged-in user with Duo Agent Platform access run commands on the gateway under certain conditions, GitLab  said in an advisory . The gateway is the service that connects a GitLab instance to AI models, and only organizations that host their own gateway need to act. The flaw is fixed in gateway versions 19.2.4, 19.3.2, and 19.4.1. The flaw is tracked as  CVE-2026-90970 . GitLab disclosed it on October 2 and rated it critical, with a CVSS score of 9.9 out of 10. GitLab runs AI Gateways for its customers and has already fixed them. Customers on GitLab.com, GitLab Dedicated, and self-managed instances that use a GitLab-hosted gateway do not need to act, the company said. Self-managed customers can instead  host their own gateway , an option GitLab offers for keeping AI request and response data inside the customer's own environment. GitLab strongly recommends that those customers update immediately. It sent that gui...

OpenAI has parted ways with three members of its safety team after they leaked private information in violation of company policies, The Wall Street Journal reported . "We have parted ways with three individuals for violating our policies on accessing and handling sensitive company information," a spokesperson for the company was quoted as saying. "Our investigation confirmed that these individuals mishandled sensitive information outside established company procedures, violating our policies and breaking the trust essential to our work." The impacted employees are Jasmine Wang , Tomek Korbak , and Mikita Balesni , the Journal reported, citing people familiar with the matter. The three researchers have all previously expressed concerns about the pace of artificial intelligence (AI) development. It's said that the individuals shared confidential information with a third-party AI-safety organization. The name of the organization was not disclosed. According...

The quarterly board meeting is two weeks out. The security team is pulling exports from the identity provider, the cloud posture tool, the vulnerability scanner, the SIEM and the EDR console. Someone is building a spreadsheet to reconcile them. Someone else is turning that spreadsheet into slides. Then a board member asks three questions: How secure is the organization, overall? What is the actual financial exposure? Is the security posture better than it was last quarter? Most security leaders cannot answer any of them with confidence. Not because the data doesn't exist, but because it lives in a dozen tools that don't share context. A new guide to confident board reporting for CISOs takes on exactly this problem. This article walks through why traditional reporting fails and what a better model looks like. Boards Have Stopped Trusting Activity Metrics For years, security reporting has run on counts. Vulnerabilities found. Patches applied. Alerts closed. P...

This week, the useful words are boring ones: inspect, cache, compile, store, trust. Each sounds harmless. Each can become an attack path when a system does a little more than people expect. A model check can run code. A cache can mix up requests. A public secret can stay useful for years. That is the lesson running through the list. Attackers do not always need a brilliant new trick. They can hide commands in public infrastructure, reuse old flaws, abuse weak defaults, or let automation stitch together a rough path that still works. Faster tools are changing the pace, but basic mistakes are still doing plenty of the work. So the interesting question this week is not “what broke?” It is “what did we assume was safe because it looked ordinary?” The full list has answers. The threats change every week. Subscribe, and we’ll alert you when each new ThreatsDay Bulletin is out.

Every security leader at a bank, insurer, or asset manager has had a version of this conversation: Security wants to eliminate a class of vulnerabilities. Engineering explains what it would take to upgrade the platform where they live. Somebody prices out the regression testing. Somebody else raises the change-freeze calendar. The finding gets an exception, a compensating control, and a date eighteen months out on the roadmap to address it. Nobody in that conversation is being unreasonable. Financial services carry more legacy software than almost any other industry for a few reasons: decades of accumulated infrastructure, regulatory obligations that reward stability, and applications where an hour of downtime is unacceptable. In that environment, minimizing change is risk management. Every dependency bump, every base image swap, every migration is a chance to break something that clears trades or moves money. So the instinct to stick to the status quo has been sound. The problem...

OpenAI on Wednesday said it identified and disrupted a coordinated distillation campaign that was designed to illicitly extract protected reasoning from its artificial intelligence (AI) models. A "core cluster of the activity," going back to the first week of July, has been attributed to individuals associated with Moonshot AI, a Chinese AI company based in Beijing. It did not cite any technical evidence to back this assessment, likely owing to security reasons. "The operators did not break our encryption, compromise a database, or gain direct access to stored user conversations," OpenAI said . "Instead, they manipulated model interactions so that protected reasoning could be reproduced in forms visible to the requester in a coordinated, scaled manner that violated our terms of service." The activity is said to have begun on July 1, 2026, initially at a low volume before it spiked on July 24 and 25, 2026, to 16,000 attempted requests using a relevan...

Google on Wednesday announced its latest frontier artificial intelligence (AI) model, Gemini 4 Argon , that it said is being rolled out to a set of trusted cyber defenders through its Fairwind Program. "It delivers frontier performance in complex workflows across real-world software engineering, enterprise knowledge work like legal and finance, and cybersecurity defense," Koray Kavukcuoglu, senior vice president of Google DeepMind and Chief AI Architect at Google, said . The development comes nearly a month after the tech giant unveiled Gemini 3.8 Flash Cyber , which it described as the most capable cybersecurity model. Like similar models from rivals Anthropic and OpenAI, Argon is assessed to be highly capable at autonomously finding, validating, and patching critical software vulnerabilities. This includes a previously unknown critical vulnerability exposing sensitive personal information across healthcare software used by hospitals worldwide. Google did not reveal...

Threat actors are abusing ChatGPT Custom GPTs to disguise them as legitimate product offerings and direct unsuspecting victims to malicious sites that employ ClickFix lures to deliver malware. Huntress, which observed the activity in late September 2026, said it marks the abuse of yet another feature in trusted artificial intelligence (AI) platforms. Prior campaigns have weaponized shared conversations with AI chatbots and malicious Claude Artifacts to distribute stealer malware and remote access trojans (RATs). Custom GPTs refer to a personalized version of ChatGPT that, as the name implies, allows users to define custom instructions, upload reference files, and enable specific skills to handle unique tasks without any coding. They are hosted on the legitimate ChatGPT website with the Custom GPT name at the top. "In the incidents we saw, victims interacted with an attacker-created Custom GPT, which was programmed to respond to their prompts with a message that includ...

AI coding agents asked to share screenshots of code changes for review have put internal company images in public GitHub repositories, security company Glow said. Its researchers found more than 13,000 internal images from developers at over 300 organizations, including customer billing records and screens of features not yet released. In most cases, they sat under developers' personal accounts, where anyone could download them but company security teams did not see them. The affected organizations include one of the world's largest tech companies, a leading AI lab, a major enterprise software provider, and a Fortune 500 travel company. Glow began contacting them on September 9,  published its findings  on September 29, and says others are likely affected too. In one case, a developer at a manufacturer with more than 100,000 employees asked an agent to check a fix to an internal billing screen. The agent created a public repository in the developer's personal GitHub ...

A malicious MCP server could trick an application built on the official  MCP Python SDK  into handing over the OAuth credentials it uses to log in to a real service, the SDK's maintainers said in a security advisory. Affected versions sent the client secret, the authorization code, and the PKCE proof key to a token endpoint the attacker controlled. The fix is in versions 1.30.0 and 2.2.0. The Model Context Protocol (MCP) is an open standard for connecting AI applications to outside tools and data, and this package is its official Python SDK for building MCP servers and clients. With the stolen credentials, the attacker can request a valid access token from the real login service. Cycode, the security firm that reported the flaw , demonstrated that full exchange in a test and says the resulting token carries whatever permissions the app was granted. The client secret is long-lived, so it keeps working until it is changed. The flaw is rated high (7.5) for the two prov...

OpenAI on Monday shelved plans to release GPT-6.1 Astra, a next-generation artificial intelligence (AI) model that was planned for an October launch, after it failed internal safety and alignment audits. The development was first reported by The Wall Street Journal. The move "marks a rare case of a major AI developer ditching a new release because of safety concerns," the news publication said. The ChatGPT maker said it made the decision to scrap its GPT-6.1 Astra model release after testing raised questions about whether it can follow user instructions without deviating from expected behavior. The Journal reported that the model exhibited higher levels of deception than its predecessor during evaluation, and failed to disclose what actions it had carried out. In some cases, it went ahead without seeking permission or attempted to use outside tools in scenarios where doing so could be deemed unsafe. "While (GPT-6.1 Astra) improved on axes such as model laziness...

OpenAI said it has made the decision to pause training of its most powerful models after one of its agents during reinforcement learning (RL) training contacted an external chatbot by exploiting a loophole in its internet-access restrictions. "An agent attempting to complete a search-based training task queried a public chatbot service through a gap in our internet-access restrictions: insufficient DNS filtering in its training sandbox," OpenAI said . "Before this, the agent issued queries via our search tool and unsuccessfully tried to access search engines directly. Note that all internet access apart from the DNS resolver in this report hit our offline webcache and therefore did not access the live internet." OpenAI said it has since added blocking controls at two independent layers to prevent this access in the first place. It also said its misalignment monitoring system detected the behavior within 15 minutes and it was acknowledged by a human reviewer...

What is IAM for AI agents? AI agents authenticate, invoke tools, and act across enterprise systems with delegated authority. IAM for AI Agents is the identity-control architecture that governs those actors. This guide covers the limits of conventional provisioning, the components that matter, how to evaluate framework choices, and what runtime evidence proves an agent behaved as intended. Identity and access management (IAM) for AI agents treats each agent as a non-human identity with a human owner, a defined purpose, scoped authorization, an expiration, and continuous monitoring. The complication is architectural. IAM platforms express intended access, while applications and infrastructure reveal what the agent actually executed. Between the two sits identity dark matter: the agents, credentials, application-local accounts, and authentication paths that central identity data never reports. A framework that cannot observe that surface produces policy intent, not assurance. Wh...