Find what you’re looking for
Browse everything
6,877 results in Threat Reports
Anthropic MCP Git Server Vulnerability: AI Supply Chain at Risk (2026)
In January 2026, three critical vulnerabilities were disclosed in Anthropic's official Model Context Protocol (MCP) Git server, a tool widely used for managing Git repositories programmatically via large language models (LLMs). The flaws—identified as CVE-2025-68143, CVE-2025-68144, and CVE-2025-68145—included two path traversal and one argument injection vulnerabilities. Attackers could exploit these bugs through prompt injection via malicious content (e.g., README files, poisoned issue descriptions) to access or overwrite arbitrary files and achieve remote code execution, without needing direct access to victim systems. Chained exploitation with Filesystem MCP enabled full system compromise until patched in late 2025. This incident highlights growing risks at the intersection of AI-driven development, supply chain dependencies, and insufficient input validation within reference implementations. As organizations increasingly rely on open-source AI tooling, securing both the application layer and its automation becomes vital in defending against evolving prompt-based and supply chain exploitation.
North Korea Supply Chain Attack Exploits VS Code Projects – 2026 Analysis
In January 2026, cybersecurity researchers uncovered a sophisticated supply chain attack targeting software developers via malicious Visual Studio Code (VS Code) projects. Threat actors linked to North Korea's Contagious Interview campaign distributed weaponized VS Code samples to compromise developer endpoints and install covert backdoors. Once inside victims' systems, the attackers could move laterally, exfiltrate sensitive source code, and access development infrastructure, risking intellectual property, customer data, and supply chain integrity. The campaign represents an evolution of social engineering tactics and demonstrates the attackers’ focus on high-leverage targets within the tech sector. This incident is highly relevant as it underscores the growing trend of software supply chain attacks leveraging development environments and trusted open-source platforms. Organizations must now reassess third-party code risks and developer security, as attackers increasingly exploit toolchains and social-engineering techniques instead of perimeter defenses.
Chainlit AI Framework Under Fire: 2024 Vulnerabilities Expose Multicloud Risk
In 2024, critical vulnerabilities were uncovered in the Chainlit open-source AI chatbot framework, enabling attackers to exploit flaws in authentication and traffic encryption protections. Malicious actors could intercept unencrypted traffic, manipulate east-west communications, and abuse privileged access across multicloud deployments, potentially leading to data exfiltration or advanced lateral movement within enterprise environments. The incident highlighted how insecure defaults and lack of robust segmentation in AI frameworks expose companies to elevated risk, demanding urgent attention from organizations relying on these technologies for customer-facing or sensitive operations. This breach underscores a broader trend of threat actors targeting open-source AI platforms, taking advantage of immature security practices inherent to many machine learning deployments. As regulatory scrutiny and supply chain attacks surge, securing AI development and deployment pipelines is increasingly essential to prevent business disruption or compliance violations.
Schneider Electric Foxboro DCS: Intel Side-Channel Flaw Threatens Critical Infrastructure (2026)
In late 2025 and early 2026, Schneider Electric disclosed a side-channel vulnerability (CVE-2018-12130) impacting its EcoStruxure Foxboro DCS product line, widely used in critical infrastructure globally. The issue, originating from Intel processor flaws, could allow authenticated local attackers to extract sensitive data via side-channel methods, potentially leading to unauthorized disclosure or manipulation of system functions. The exploit primarily affects specific Foxboro DCS servers and workstations running on vulnerable Intel CPUs. Schneider Electric issued upgrades and remediation guidance while urging organizations to implement defense-in-depth strategies to mitigate risk. This incident highlights ongoing industry concerns over hardware-level vulnerabilities affecting operational technology in high-stakes sectors such as energy and manufacturing. As threat actors increasingly target supply chain and embedded flaws, organizations are under mounting pressure from regulators and customers to update aging infrastructure, strengthen segmentation, and accelerate threat detection capabilities.
Rockwell Automation Verve Asset Manager Vulnerabilities: 2026 Lessons for Critical Infrastructure
In January 2026, Rockwell Automation disclosed two significant vulnerabilities (CVE-2025-14376, CVE-2025-14377) in its Verve Asset Manager product. These flaws were rooted in insecure and cleartext storage of sensitive data within the legacy ADI server and Ansible playbook components, impacting versions 1.33 through 1.41.3. Exploitation could have allowed attackers with system or network access to retrieve confidential data from environment variables and process files, potentially facilitating lateral movement or further compromises. The issues were addressed in version 1.42, and vulnerable components were made optional in newer releases. This incident is particularly relevant amid heightened attention to supply chain risk and critical infrastructure cybersecurity. As industrial control vendors face rising regulatory pressure and expansion of zero-trust mandates, unencrypted data storage flaws highlight the urgent need for comprehensive data-in-transit and at-rest protections.
Critical RCE Flaws in Microsoft & Anthropic MCP Servers Expose AI Workloads to Cloud Takeover
In early June 2024, researchers disclosed critical remote code execution (RCE) vulnerabilities in the Model Context Protocol (MCP) servers operated by Microsoft and Anthropic, exposing integral AI infrastructure to severe cloud takeover risks. Threat actors could leverage these flaws to execute arbitrary code and potentially gain privileged access to sensitive data or underlying cloud platforms, risking widespread lateral movement and data exfiltration. The vulnerabilities arose from insufficient encryption and a lack of east-west segmentation controls, enabling potential attackers to compromise interconnected AI workloads and services. Both vendors moved quickly to deploy patches, but the initial exposure window highlighted deep-seated risks in shared AI service architectures. This incident highlights growing adversary focus on AI model supply chains and service interconnects, with attackers exploiting new protocol vulnerabilities. Rising adoption of AI-driven services across industries has amplified attack surfaces, urging enterprises to enhance zero trust segmentation, encrypted traffic controls, and multicloud observability to meet evolving compliance and operational challenges.
Google Gemini Exploited: Calendar Invites Become AI Attack Vector in 2024
In early 2024, researchers identified a critical indirect prompt injection vulnerability affecting Google Gemini, Google's flagship AI suite. Attackers exploited calendar invites as a covert vector to manipulate Gemini's AI context, successfully bypassing native privacy filters and accessing sensitive user data. The attack leveraged the insertion of malicious prompts within innocuous-looking calendar items, which Gemini processed automatically, leading to unauthorized access and data exposure. Google responded by issuing incremental updates, but the incident highlighted inherent risks in automated AI integrations with productivity platforms reliant on user-generated content. This incident underscores the growing prevalence of AI/ML abuse through indirect attack vectors such as prompt injection. With threat actors increasingly targeting large language models in enterprise environments, security programs must rapidly adapt to cover AI-specific exposures, ensure robust segmentation, and incorporate real-time anomaly detection to defend against evolving risks.
Punycode Phishing: How IDN Abuse Enabled Stealth Attacks in 2026
In January 2026, security researchers identified a surge in phishing attacks leveraging Internationalized Domain Names (IDNs) encoded with Punycode, enabling attackers to create visually deceptive domains that closely resemble legitimate ones. By substituting standard ASCII characters with similar-looking Unicode characters, threat actors bypassed traditional detection, tricking users into visiting fraudulent sites and unknowingly exposing credentials or sensitive information. The attack was uncovered through DNS log analysis, revealing repeated internal access attempts to encoded domains such as xn--yutube-wqf.com, demonstrating the sophistication and stealth of this social engineering tactic. This incident highlights the growing prevalence of advanced phishing campaigns using homoglyph attacks and encoded domains, underscoring the need for updated detection routines and user awareness. Organizations face increased operational risk as threat actors exploit gaps caused by internationalization and encoding, making timely monitoring and DNS log analysis crucial.
Predator Bots Hit APIs at Scale: Are Your Defenses Ready for Autonomous Threats?
In 2024, a surge in highly sophisticated "predator bots" led to a large-scale wave of automated attacks targeting critical APIs across multiple industries. Threat actors leveraged AI-driven automation to mimic human behavior and evade detection, resulting in credential theft, fraudulent transactions, account takeovers, and widespread data scraping. The attacks exploited shadow APIs and business logic flaws, bypassed traditional defenses like IP filtering and CAPTCHAs, and drained significant revenue through scalping and abuse. The economic impact has been severe, with business losses estimated in the billions, and operational teams scrambling to regain visibility and control. This incident highlights the urgent need for modern, adaptive security that operates at machine speed. As attackers weaponize automation and AI, organizations face mounting pressure to implement layered defenses, proactive anomaly detection, and API-centric protection to prevent sophisticated fraud, data loss, and erosion of customer trust.
Ingram Micro 2025 Ransomware Breach: Over 42,000 Impacted in Supply Chain Attack
In July 2025, information technology distributor Ingram Micro experienced a significant ransomware attack that compromised its internal systems and resulted in the unauthorized access and potential theft of sensitive personal information for over 42,000 individuals. Attackers infiltrated the company’s network, deployed ransomware, and encrypted critical data, causing temporary disruption to business operations. Ingram Micro responded by shutting down affected systems, launching an investigation, and notifying impacted individuals, highlighting gaps in east-west traffic security and incident detection. This breach underscores the ongoing resurgence of ransomware attacks targeting supply chain companies and IT providers. With attackers refining lateral movement techniques, organizations face mounting pressure to adopt robust segmentation, real-time anomaly detection, and comprehensive data protection strategies in compliance with evolving regulatory expectations.
Jordanian Access Broker Case Exposes Credential Sales Across 50 Enterprises
In May 2023, law enforcement identified Feras Khalil Ahmad Albashiti—a Jordanian national—operating as an initial access broker, selling unauthorized access to over 50 corporate networks via an underground cybercrime forum. Acting under the handle "r1z," Albashiti exchanged credentials for cryptocurrency on at least one occasion with an undercover officer, exposing illicit sales tied to fraud and abuse of privileged network access. Arrested in Georgia and extradited to the US in July 2024, he pleaded guilty to charges of fraud involving access credentials. Sentencing is scheduled for May 2026, with potential penalties of up to 10 years imprisonment and substantial fines. This case highlights the increasingly organized role of initial access brokers in cybercrime, where privileged access is sold to facilitate ransomware, espionage, and data theft. The incident underscores ongoing risks posed by the thriving market for stolen credentials used to compromise enterprise environments.
Supreme Court and Agency Data Breached via Stolen Credentials: The 2023 Instagram Leak
In late 2023, a Tennessee man illicitly accessed the U.S. Supreme Court’s restricted electronic filing system, as well as accounts at AmeriCorps and the Department of Veterans Affairs, through the repeated use of stolen credentials. Over multiple months, Nicholas Moore gained unauthorized entry to sensitive government systems at least 25 times, collecting and exfiltrating personal, legal, and health data. He then publicized this sensitive information via his Instagram handle, @ihackedthegovernment, exposing government, AmeriCorps, and veteran data, including personal identifiers and privileged health information. This case underscores a rise in breaches involving compromised credentials, lateral movement, and public boasting on social media. With persistent attacker focus on governmental targets and data exfiltration, it exemplifies the ongoing risks of inadequate east-west and data-in-transit security, as well as the compliance pressure on federal agencies to shore up access controls and insider threat monitoring.