✨ No need to do from scratch. Deploy a Validated Containment Architecture built for your AI platform. →Deploy a Validated Containment Architecture for your AI platform. →A Validated Containment Architecture for your AI platform. →Validated Containment Architectures are here. →Contain Threats NowExplore✨
Computer Software/Engineering
Breach intelligence, attack campaigns, and threat reports targeting the Computer Software/Engineering sector.
Explore Other Sectors
Computer Software/Engineering Threat Reports
Malicious AI Extensions in VSCode Marketplace Steal Developer Data, Impacting 1.5 Million Users
In January 2026, researchers at Koi Security discovered that two AI-powered Visual Studio Code (VSCode) Marketplace extensions—ChatGPT – 中文版 and ChatMoss (CodeMoss)—were secretly exfiltrating developer files and sensitive data to China-based servers. Together, these malicious extensions had been installed 1.5 million times and collected data using real-time file monitoring, workspace file harvesting, and covert user profiling via embedded commercial analytics SDKs. The compromised extensions transmitted not only source code but potentially included API keys, configuration, and credential files without user consent, representing a major supply-chain compromise in the software development ecosystem. This incident highlights the persistent risks developers face from supply chain attacks through third-party plugins. As AI-driven code assistants surge in popularity, attackers are increasingly exploiting trusted extension marketplaces to deploy sophisticated data-stealing campaigns, raising urgent concerns for software security, compliance, and marketplace governance.
6 months ago
Kill Chain
AI-Generated Code Exposes New Honeypot Security Risks at Intruder (2026)
In January 2026, Intruder Security revealed an application security vulnerability in their intentionally vulnerable honeypot, stemming from AI-generated code that mishandled client-supplied IP headers. The AI-assisted system incorrectly trusted IP values in HTTP headers without enforcing a trust boundary, allowing attackers to inject payloads or spoof source IP information. This oversight, undetected by common static analysis tools, resulted in attacker-controlled inputs influencing system logic, posing potential risks for local file disclosure or server-side request forgery had the vulnerable code path been used differently. While the actual impact remained low due to the isolated nature of the honeypot, the incident highlights significant gaps in current AI-assisted development and security review processes. This event underscores the growing prevalence of AI-generated vulnerabilities and the limitations of automated security tools in identifying nuanced flaws. As enterprises increasingly rely on AI-driven coding and automation, such oversights are likely to become more common, emphasizing the need for robust validation and updated security governance.
6 months ago
Kill Chain
Pwn2Own Automotive 2026: 29 Zero-Days Unmasked in Next-Gen Vehicle Tech
At Pwn2Own Automotive 2026 in Tokyo, security researchers exploited 29 zero-day vulnerabilities on the second day alone, targeting fully patched electric vehicle (EV) chargers, in-vehicle infotainment systems (IVI), and automotive operating systems. Over $439,000 in prizes was awarded for these findings, impacting vendors like Phoenix Contact, ChargePoint, Grizzl-E, Kenwood, Alpine, and Alpitronic HYC50. Notably, researcher teams used advanced exploit chains to achieve root access, demonstrating how attackers could potentially compromise critical automotive technology with little or no prior warning. The event underscores the growing attack surface in connected vehicles and the persistent risk posed by unreported vulnerabilities. This incident highlights a significant surge in automotive cybersecurity threats, especially as EV and smart vehicle adoption accelerates. The frequency and sophistication of these exploits illustrate the urgency for automakers and suppliers to prioritize vulnerability management, rapid patch deployment, and layered defense strategies to protect both consumer safety and data integrity.
6 months ago
Kill Chain
INC Ransomware OpSec Fail Uncovers Data from 12 U.S. Organizations
In January 2026, an operational security lapse in the INC ransomware group's infrastructure enabled Cyber Centaurs researchers to recover encrypted data exfiltrated from twelve U.S. organizations. The investigation began after a RainINC ransomware attack on a client’s production SQL Server. Forensic analysis traced renamed binaries, PowerShell scripts, and usage of the Restic backup tool, revealing attacker scripts with hardcoded credentials and references to persistent cloud storage. By enumerating the attacker-controlled repositories, researchers identified encrypted data from healthcare, manufacturing, technology, and services firms, then decrypted and preserved it in coordination with law enforcement. This case highlights a rare opportunity where attacker mistakes allowed post-breach data retrieval for unrelated victim organizations. The incident underscores a growing trend in ransomware operations leveraging legitimate backup and exfiltration tools, persistent attacker infrastructure, and the importance of thorough incident response for uncovering wider impacts.
6 months ago
Kill Chain
GitLab Faces Critical 2FA Bypass and DoS Vulnerabilities in 2026
In January 2026, GitLab urgently patched several high-severity vulnerabilities affecting its widely used Community and Enterprise Editions. The most critical issue, tracked as CVE-2026-0723, allowed attackers with knowledge of a user's account ID to bypass two-factor authentication controls by submitting forged device responses, resulting from unchecked return values in authentication services. In addition, GitLab addressed multiple denial-of-service (DoS) vulnerabilities, including CVE-2025-13927 and CVE-2025-13928, which potentially let unauthenticated threat actors trigger service outages through malformed authentication data and improper API endpoint authorization checks. Immediate patches were released to mitigate the risks of account takeover, service disruption, and operational downtime across a user base spanning major enterprises and nearly 6,000 exposed internet-facing instances. This breach stands out in the context of rising attacks exploiting authentication weaknesses and API logic flaws across the software supply chain. As critical open-source DevSecOps platforms like GitLab underpin enterprise workflows, attackers increasingly target authentication and availability gaps, aligning with regulatory scrutiny and the growing demand for robust zero trust controls.
6 months ago
Kill Chain
AI-Powered Android Malware Unleashes New Click-Fraud Wave via Xiaomi App Store
In January 2026, cybersecurity researchers at Dr.Web uncovered a sophisticated new Android malware family distributed via Xiaomi’s GetApps, popular third-party APK sites, and messaging platforms like Telegram and Discord. This malware leverages AI-driven image analysis using Google’s TensorFlow.js to identify and autonomously click on hidden browser ads within compromised apps, particularly games, simulating user behavior without obvious signs to victims. The malware is delivered through legitimate-looking apps, which update with malicious payloads post-installation. Impacts include increased battery consumption, higher data charges, and indirect monetization for attackers. This incident exemplifies the evolution of mobile ad fraud TTPs, as attackers increasingly deploy AI/ML for advanced automation and evasion. The trend signals rising risks to mobile advertising integrity and higher scrutiny for app stores’ vetting processes, especially on third-party and OEM-specific app markets.
6 months ago
Kill Chain
CERT/CC Alert: binary-parser npm Vulnerability Puts Node.js Apps at Risk
In November 2025, a critical vulnerability (CVE-2026-1245) was disclosed in the widely used binary-parser npm library, enabling attackers to execute arbitrary JavaScript code on impacted Node.js applications. The issue stemmed from unsanitized user-supplied values in dynamically generated parser code, leaving systems relying on untrusted parser definitions open to privilege-level code execution and potential compromise of local data, application logic, or even execution of system commands. CERT/CC publicly warned about this supply-chain risk, urging organizations to upgrade to binary-parser v2.3.0 and avoid processing untrusted parser configurations. This incident is a stark reminder of the supply-chain risks inherent in open-source dependencies, especially those that permit dynamic code generation. As exploitation of package vulnerabilities continues to rise, regulators and CISOs are placing increasing importance on proactive dependency management and runtime validation in development and DevOps pipelines.
6 months ago
Kill Chain
Chainlit 2026 Supply-Chain Flaws Let Hackers Breach Cloud AI Environments
In early 2026, two high-severity vulnerabilities ('ChainLeak') were discovered in Chainlit, a widely adopted open-source conversational AI framework, exposing cloud environments to significant risk. The flaws—CVE-2026-22218 (arbitrary file read) and CVE-2026-22219 (server-side request forgery)—could be exploited without user interaction, allowing attackers to access sensitive files and internal services on internet-facing production systems. Zafran Labs demonstrated that chaining both vulnerabilities enabled full-system compromise and lateral movement within enterprise cloud environments before a patch (v2.9.4) was released in December 2025. This incident highlights a growing threat vector in the AI software supply chain, especially as critical business and academic applications increasingly rely on rapidly evolving open-source frameworks. With adversaries targeting common components and fast-moving cloud deployments, organizations face pressure to update promptly and re-evaluate their security posture against similar zero-day and supply-chain risks.
6 months ago
Kill Chain
Credential Stuffing Attempt at PcComponentes: 2024 Incident Overview
In June 2024, Spanish online retailer PcComponentes confirmed that its systems were targeted by a large-scale credential stuffing attack. While the company denied reports of a data breach affecting 16 million customers, it acknowledged that threat actors attempted to use previously leaked credentials to gain unauthorized access to customer accounts. No evidence of infrastructure compromise or mass data exfiltration was found, and PcComponentes’ internal investigation revealed that protective measures limited the attack’s impact. This incident highlights the ongoing challenges facing retailers from credential-based attacks, emphasizing the importance of stronger identity and access controls. The surge in credential stuffing campaigns reflects broader trends in attacker automation and customer credential reuse across online services.
6 months ago
Kill Chain
VoidLink: The First AI-Generated Linux Malware Framework Exposes New Cybersecurity Risks
In late 2025, cybersecurity researchers uncovered VoidLink, a sophisticated Linux malware framework reportedly developed primarily by a single actor leveraging artificial intelligence. Analyses by Check Point Research and Sysdig identified operational artifacts and systematic code features—like consistent formatting, template-based responses, and AI-generated debug logs—suggesting heavy use of large language models in the malware’s rapid development. The threat actor, operating in a Chinese-language environment, utilized AI tools such as TRAE SOLO to expedite Spec Driven Development, producing 88,000 lines of attack-ready code in less than a week. While no real-world infections have been reported to date, VoidLink’s capabilities are significant, aiming at stealthy, long-term access to Linux-based cloud infrastructure. The emergence of AI-generated attack frameworks like VoidLink signals a pivotal change in cybercrime. Advanced threat creation, once limited to state actors or skilled teams, can now be accomplished rapidly by individuals utilizing off-the-shelf AI tools. This trend increases the urgency for organizations to adapt AI-enhanced defenses and revisit zero trust, segmentation, and cloud security controls.
6 months ago
Kill Chain
Zoom & GitLab Issue Critical Security Patches for RCE, DoS, and 2FA Bypass—January 2026
In January 2026, Zoom and GitLab simultaneously released critical security updates to mitigate multiple high-severity vulnerabilities uncovered in their respective platforms. Zoom's most impactful flaw, CVE-2026-22844 (CVSS 9.9), affected its Node Multimedia Routers (MMRs), potentially enabling a meeting participant to perform remote code execution via a command injection. GitLab, meanwhile, addressed several vulnerabilities including two denial-of-service (DoS) issues and a two-factor authentication (2FA) bypass flaw (CVE-2026-0723), which could let malicious actors disrupt services or compromise user accounts if they knew credential IDs. While no active exploitation was reported, organizations using Zoom’s Node MMR module and GitLab’s CE/EE deployments were urged to patch immediately to avoid significant business disruption or data compromise. The disclosure of these vulnerabilities highlights the increasing sophistication and severity of threats targeting software supply chains and critical collaboration platforms. With attackers frequently seeking novel vectors for RCE, DoS, and authentication bypass, timely patching and robust segmentation remain crucial.
6 months ago
Kill Chain
Chainlit AI Framework Flaws Expose Sensitive Data: What Organizations Need to Know
In January 2026, serious security flaws were disclosed in the popular open-source Chainlit artificial intelligence (AI) framework, exposing organizations to substantial data theft and privilege escalation risks. The vulnerabilities—CVE-2026-22218 (arbitrary file read) and CVE-2026-22219 (server-side request forgery/SSRF)—allowed authenticated attackers to steal sensitive files, such as API keys and credentials, and leverage SSRF to access internal or cloud network services. Exploiting these flaws, attackers could combine vectors for lateral movement across environments, potentially leading to broader compromise of AI-powered systems and cloud infrastructures. Both issues were patched in version 2.9.4 following responsible disclosure. This breach is highly relevant as adoption of AI frameworks accelerates and attackers increasingly target embedded infrastructure weaknesses, combining known vulnerabilities with emerging AI application risks. The incident underscores why organizations must adapt security controls and continuously monitor new technology stacks for widely exploitable flaws.
6 months ago
Kill Chain
Stop Active Cloud Data Exfiltration
Aviatrix Breach Lock helps teams instantly identify what data is leaving the environment, from which workload, and where it’s going — during an active breach.
Looking for threats in a different sector?
Browse All Threat Reports