Vulnerability | Breaking Cybersecurity News | The Hacker News

Cybersecurity researchers have uncovered critical remote code execution vulnerabilities impacting major artificial intelligence (AI) inference engines, including those from Meta, Nvidia, Microsoft, and open-source PyTorch projects such as vLLM and SGLang. "These vulnerabilities all traced back to the same root cause: the overlooked unsafe use of ZeroMQ (ZMQ) and Python's pickle deserialization," Oligo Security researcher Avi Lumelsky said in a report published Thursday. At its core, the issue stems from what has been described as a pattern called ShadowMQ , in which the insecure deserialization logic has propagated to several projects as a result of code reuse. The root cause is a vulnerability in Meta's Llama large language model (LLM) framework ( CVE-2024-50050 , CVSS score: 6.3/9.3) that was patched by the company last October. Specifically, it involved the use of ZeroMQ's recv_pyobj() method to deserialize incoming data using Python's pickle module. ...

Cybersecurity researchers are sounding the alert about an authentication bypass vulnerability in Fortinet Fortiweb WAF that could allow an attacker to take over admin accounts and completely compromise a device. "The watchTowr team is seeing active, indiscriminate in-the-wild exploitation of what appears to be a silently patched vulnerability in Fortinet's FortiWeb product," Benjamin Harris, watchTowr CEO and founder, said in a statement. "Patched in version 8.0.2 , the vulnerability allows attackers to perform actions as a privileged user - with in-the-wild exploitation focusing on adding a new administrator account as a basic persistence mechanism for the attackers." The cybersecurity company said it was able to successfully reproduce the vulnerability and create a working proof-of-concept (Poc). It has also released an artifact generator tool for the authentication bypass to help identify susceptible devices. According to details shared by Defused a...

The U.S. Cybersecurity and Infrastructure Security Agency (CISA) on Wednesday added a critical security flaw impacting WatchGuard Fireware to its Known Exploited Vulnerabilities ( KEV ) catalog, based on evidence of active exploitation. The vulnerability in question is CVE-2025-9242 (CVSS score: 9.3), an out-of-bounds write vulnerability affecting Fireware OS 11.10.2 up to and including 11.12.4_Update1, 12.0 up to and including 12.11.3 and 2025.1. It was patched by WatchGuard in September. "WatchGuard Firebox contains an out-of-bounds write vulnerability in the OS iked process that may allow a remote unauthenticated attacker to execute arbitrary code," CISA said in an advisory. Details of the vulnerability were shared by watchTowr Labs last month, with the cybersecurity company stating that the issue stems from a missing length check on an identification buffer used during the IKE handshake process. "The server does attempt certificate validation, but that valid...

Amazon's threat intelligence team on Wednesday disclosed that it observed an advanced threat actor exploiting two then-zero-day security flaws in Cisco Identity Service Engine (ISE) and Citrix NetScaler ADC products as part of attacks designed to deliver custom malware. "This discovery highlights the trend of threat actors focusing on critical identity and network access control infrastructure – the systems enterprises rely on to enforce security policies and manage authentication across their networks," CJ Moses, CISO of Amazon Integrated Security, said in a report shared with The Hacker News. The attacks were flagged by its MadPot honeypot network, with the activity weaponizing the following two vulnerabilities - CVE-2025-5777 or Citrix Bleed 2 (CVSS score: 9.3) - An insufficient input validation vulnerability in Citrix NetScaler ADC and Gateway that could be exploited by an attacker to bypass authentication. (Fixed by Citrix in June 2025 ) CVE-2025-20337 (CV...

Microsoft on Tuesday released patches for 63 new security vulnerabilities identified in its software, including one that has come under active exploitation in the wild. Of the 63 flaws, four are rated Critical and 59 are rated Important in severity. Twenty-nine of these vulnerabilities are related to privilege escalation, followed by 16 remote code execution, 11 information disclosure, three denial-of-service (DoS), two security feature bypass, and two spoofing bugs. The patches are in addition to the 27 vulnerabilities the Windows maker addressed in its Chromium-based Edge browser since the release of October 2025's Patch Tuesday update. The zero-day vulnerability that has been listed as exploited in Tuesday's update is CVE-2025-62215 (CVSS score: 7.0), a privilege escalation flaw in Windows Kernel. The Microsoft Threat Intelligence Center (MSTIC) and Microsoft Security Response Center (MSRC) have been credited with discovering and reporting the issue. "Concurre...

Google's Mandiant Threat Defense on Monday said it discovered n-day exploitation of a now-patched security flaw in Gladinet's Triofox file-sharing and remote access platform. The critical vulnerability, tracked as CVE-2025-12480 (CVSS score: 9.1), allows an attacker to bypass authentication and access the configuration pages, resulting in the upload and execution of arbitrary payloads.  The tech giant said it observed a threat cluster tracked as UNC6485 weaponizing the flaw as far back as August 24, 2025, nearly a month after Gladinet released patches for the flaw in version 16.7.10368.56560 . It's worth noting that CVE-2025-12480 is the third flaw in Triofox that has come under active exploitation this year alone, after CVE-2025-30406 and CVE-2025-11371 . "Added protection for the initial configuration pages," according to release notes for the software. "These pages can no longer be accessed after Triofox has been set up." Mandiant said the th...

A China-linked threat actor has been attributed to a cyber attack targeting an U.S. non-profit organization with an aim to establish long-term persistence, as part of broader activity aimed at U.S. entities that are linked to or involved in policy issues. The organization, according to a report from Broadcom's Symantec and Carbon Black teams, is "active in attempting to influence U.S. government policy on international issues." The attackers managed to gain access to the network for several weeks in April 2025. The first sign of activity occurred on April 5, 2025, when mass scanning efforts were detected against a server by leveraging various well-known exploits, including CVE-2022-26134 (Atlassian), CVE-2021-44228 (Apache Log4j), CVE-2017-9805 (Apache Struts), and CVE-2017-17562 (GoAhead Web Server). Symantec and Carbon Black told The Hacker News that there is no indication that these exploitation efforts were successful. It's suspected that the attackers ul...

A previously unknown threat activity cluster has been observed impersonating Slovak cybersecurity company ESET as part of phishing attacks targeting Ukrainian entities. The campaign, detected in May 2025, is tracked by the security outfit under the moniker InedibleOchotense , describing it as Russia-aligned. "InedibleOchotense sent spear-phishing emails and Signal text messages, containing a link to a trojanized ESET installer, to multiple Ukrainian entities," ESET said in its APT Activity Report Q2 2025–Q3 2025 shared with The Hacker News. InedibleOchotense is assessed to share tactical overlaps with a campaign documented by EclecticIQ that involved the deployment of a backdoor called BACKORDER and by CERT-UA as UAC-0212 , which it describes as a sub-cluster within the Sandworm (aka APT44) hacking group. While the email message is written in Ukrainian, ESET said the first line uses a Russian word, likely indicating a typo or a translation error. The email, which purp...

Cisco on Wednesday disclosed that it became aware of a new attack variant that's designed to target devices running Cisco Secure Firewall Adaptive Security Appliance (ASA) Software and Cisco Secure Firewall Threat Defense (FTD) Software releases that are susceptible to CVE-2025-20333 and CVE-2025-20362 . "This attack can cause unpatched devices to unexpectedly reload, leading to denial-of-service (DoS) conditions," the company said in an updated advisory, urging customers to apply the updates as soon as possible. Both vulnerabilities were disclosed in late September 2025, but not before they were exploited as zero-day vulnerabilities in attacks delivering malware such as RayInitiator and LINE VIPER , according to the U.K. National Cyber Security Centre (NCSC). While successful exploitation of CVE-2025-20333 allows an attacker to execute arbitrary code as root using crafted HTTP requests, CVE-2025-20362 makes it possible to access a restricted URL without authentica...

Cybersecurity researchers have disclosed a new set of vulnerabilities impacting OpenAI's ChatGPT artificial intelligence (AI) chatbot that could be exploited by an attacker to steal personal information from users' memories and chat histories without their knowledge. The seven vulnerabilities and attack techniques, according to Tenable, were found in OpenAI's GPT-4o and GPT-5 models. OpenAI has since addressed some of them .  These issues expose the AI system to indirect prompt injection attacks , allowing an attacker to manipulate the expected behavior of a large language model (LLM) and trick it into performing unintended or malicious actions, security researchers Moshe Bernstein and Liv Matan said in a report shared with The Hacker News. The identified shortcomings are listed below - Indirect prompt injection vulnerability via trusted sites in Browsing Context, which involves asking ChatGPT to summarize the contents of web pages with malicious instructions added...

The U.S. Cybersecurity and Infrastructure Security Agency (CISA) on Tuesday added two security flaws impacting Gladinet and Control Web Panel (CWP) to its Known Exploited Vulnerabilities ( KEV ) catalog, citing evidence of active exploitation in the wild. The vulnerabilities in question are listed below - CVE-2025-11371 (CVSS score: 7.5) - A vulnerability in files or directories accessible to external parties in Gladinet CentreStack and Triofox that could result in unintended disclosure of system files. CVE-2025-48703 (CVSS score: 9.0) - An operating system command injection vulnerability in Control Web Panel (formerly CentOS Web Panel) that results in unauthenticated remote code execution via shell metacharacters in the t_total parameter in a filemanager changePerm request. The development comes weeks after cybersecurity company Huntress said it detected active exploitation attempts targeting CVE-2025-11371, with unknown threat actors leveraging the flaw to run reconnaissan...

Cybersecurity researchers have disclosed details of four security flaws in Microsoft Teams that could have exposed users to serious impersonation and social engineering attacks. The vulnerabilities "allowed attackers to manipulate conversations, impersonate colleagues, and exploit notifications," Check Point said in a report shared with The Hacker News. Following responsible disclosure in March 2024, some of the issues were addressed by Microsoft in August 2024 under the CVE identifier CVE-2024-38197, with subsequent patches rolled out in September 2024 and October 2025. In a nutshell, these shortcomings make it possible to alter message content without leaving the "Edited" label and sender identity and modify incoming notifications to change the apparent sender of the message, thereby allowing an attacker to trick victims into opening malicious messages by making them appear as if they are coming from a trusted source, including high-profile C-suite executives...

Google's artificial intelligence (AI)-powered cybersecurity agent called Big Sleep has been credited by Apple for discovering as many as five different security flaws in the WebKit component used in its Safari web browser that, if successfully exploited, could result in a browser crash or memory corruption. The list of vulnerabilities is as follows - CVE-2025-43429 - A buffer overflow vulnerability that may lead to an unexpected process crash when processing maliciously crafted web content (addressed through improved bounds checking) CVE-2025-43430 - An unspecified vulnerability that could result in an unexpected process crash when processing maliciously crafted web content (addressed through improved state management) CVE-2025-43431 & CVE-2025-43433 - Two unspecified vulnerabilities that may lead to memory corruption when processing maliciously crafted web content (addressed through improved memory handling) CVE-2025-43434 - A use-after-free vulnerability that may ...

The Australian Signals Directorate (ASD) has issued a bulletin about ongoing cyber attacks targeting unpatched Cisco IOS XE devices in the country with a previously undocumented implant known as BADCANDY . The activity, per the intelligence agency, involves the exploitation of CVE-2023-20198 (CVSS score: 10.0), a critical vulnerability that allows a remote, unauthenticated attacker to create an account with elevated privileges and use it to seize control of susceptible systems. The security defect has come under active exploitation in the wild since last 2023, with China-linked threat actors like Salt Typhoon weaponizing it in recent months to breach telecommunications providers. ASD noted that variations of BADCANDY have been detected since October 2023, with a fresh set of attacks continuing to be recorded in 2024 and 2025. As many as 400 devices in Australia are estimated to have been compromised with the malware since July 2025, out of which 150 devices were infected in Oct...

OpenAI has announced the launch of an "agentic security researcher" that's powered by its GPT-5 large language model (LLM) and is programmed to emulate a human expert capable of scanning, understanding, and patching code. Called Aardvark , the artificial intelligence (AI) company said the autonomous agent is designed to help developers and security teams flag and fix security vulnerabilities at scale. It's currently available in private beta. "Aardvark continuously analyzes source code repositories to identify vulnerabilities, assess exploitability, prioritize severity, and propose targeted patches," OpenAI noted . It works by embedding itself into the software development pipeline, monitoring commits and changes to codebases, detecting security issues and how they might be exploited, and proposing fixes to address them using LLM-based reasoning and tool-use. Powering the agent is GPT‑5 , which OpenAI introduced in August 2025. The company describes it...

A China-affiliated threat actor known as UNC6384 has been linked to a fresh set of attacks exploiting an unpatched Windows shortcut vulnerability to target European diplomatic and government entities between September and October 2025. The activity targeted diplomatic organizations in Hungary, Belgium, Italy, and the Netherlands, as well as government agencies in Serbia, Arctic Wolf said in a technical report published Thursday. "The attack chain begins with spear-phishing emails containing an embedded URL that is the first of several stages that lead to the delivery of malicious LNK files themed around European Commission meetings, NATO-related workshops, and multilateral diplomatic coordination events," the cybersecurity company said. The files are designed to exploit ZDI-CAN-25373 to trigger a multi-stage attack chain that culminates in the deployment of the PlugX malware using DLL side-loading. PlugX is a remote access trojan that's also referred to as Destroy...

The exploitation of a recently disclosed critical security flaw in Motex Lanscope Endpoint Manager has been attributed to a cyber espionage group known as Tick . The vulnerability, tracked as CVE-2025-61932 (CVSS score: 9.3), allows remote attackers to execute arbitrary commands with SYSTEM privileges on on-premise versions of the program. JPCERT/CC, in an alert issued this month, said that it has confirmed reports of active abuse of the security defect to drop a backdoor on compromised systems. Tick, also known as Bronze Butler, Daserf, REDBALDKNIGHT, Stalker Panda, Stalker Taurus, and Swirl Typhoon (formerly Tellurium), is a suspected Chinese cyber espionage actor known for its extensive targeting of East Asia, specifically Japan. It's assessed to be active since at least 2006. "We're aware of very targeted activity in Japan and believe the exploitation by Bronze Butler was limited to sectors aligned with their intelligence objectives," Rafe Pilling, directo...