✨ 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
Critical RCE Vulnerability in SGLang via Malicious GGUF Model Files
In April 2026, a critical vulnerability (CVE-2026-5760) was identified in SGLang, an open-source framework for serving large language models. The flaw resides in the reranking endpoint (/v1/rerank), where unsandboxed Jinja2 template rendering allows remote code execution (RCE) when processing malicious GPT-Generated Unified Format (GGUF) model files. Exploitation enables attackers to execute arbitrary code on the server, potentially leading to data exfiltration, system manipulation, or denial-of-service attacks. ([kb.cert.org](https://kb.cert.org/vuls/id/915947?utm_source=openai)) This incident underscores the importance of secure template rendering practices in AI model serving frameworks. Organizations utilizing SGLang should promptly update to a patched version and implement recommended mitigations to prevent exploitation. ([thehackernews.com](https://thehackernews.com/2026/04/sglang-cve-2026-5760-cvss-98-enables.html?utm_source=openai))
3 months ago
Kill Chain
Anthropic MCP Design Flaw Enables Remote Code Execution
In April 2026, security researchers identified a critical design flaw in Anthropic's Model Context Protocol (MCP) that enables remote code execution (RCE) across systems utilizing vulnerable MCP implementations. This systemic vulnerability affects over 7,000 publicly accessible servers and software packages with more than 150 million downloads. The flaw arises from unsafe defaults in MCP's configuration over the STDIO transport interface, allowing attackers to execute arbitrary OS commands and access sensitive data. Despite the disclosure of multiple CVEs, including CVE-2025-49596 and CVE-2026-22252, Anthropic has stated that the protocol's behavior is "expected," leaving the core issue unaddressed. ([thehackernews.com](https://thehackernews.com/2026/04/anthropic-mcp-design-vulnerability.html?utm_source=openai)) This incident underscores the escalating risks within the AI supply chain, as AI-powered integrations inadvertently expand the attack surface. Organizations are advised to implement mitigations such as blocking public IP access to sensitive services, monitoring MCP tool invocations, running MCP-enabled services in a sandbox, treating external MCP configuration input as untrusted, and installing MCP servers only from verified sources. ([thehackernews.com](https://thehackernews.com/2026/04/anthropic-mcp-design-vulnerability.html?utm_source=openai))
3 months ago
Kill Chain
Vercel's 2026 Security Breach: Lessons in Third-Party Integration Risks
In April 2026, Vercel, a prominent web infrastructure provider, experienced a security breach originating from a compromised third-party AI tool, Context.ai. An attacker exploited this vulnerability to gain unauthorized access to a Vercel employee's Google Workspace account, subsequently infiltrating Vercel's internal systems. This intrusion led to the exposure of certain environment variables not marked as 'sensitive,' potentially affecting a limited subset of customers. Vercel has since engaged incident response experts and notified law enforcement to address the situation. ([vercel.com](https://vercel.com/kb/bulletin/vercel-april-2026-security-incident?utm_source=openai)) This incident underscores the escalating risks associated with third-party integrations and the necessity for robust security measures. The breach highlights the importance of vigilant monitoring and management of OAuth applications to prevent unauthorized access and protect sensitive data.
3 months ago
Kill Chain
Vercel's April 2026 Security Breach: Lessons in Third-Party Integration Risks
In April 2026, Vercel, a prominent cloud platform, experienced a security breach originating from a compromised third-party AI tool, Context AI. An attacker exploited OAuth tokens from Context AI to access a Vercel employee's Google Workspace account, subsequently infiltrating Vercel's internal systems. This intrusion led to unauthorized access to certain customer data, including non-sensitive environment variables such as API keys and database credentials. Vercel promptly engaged external incident response experts, notified law enforcement, and advised affected customers to rotate potentially exposed credentials. The company's open-source projects, Next.js and Turbopack, were confirmed unaffected. ([techcrunch.com](https://techcrunch.com/2026/04/20/app-host-vercel-confirms-security-incident-says-customer-data-was-stolen-via-breach-at-context-ai/?utm_source=openai)) This incident underscores the escalating risks associated with third-party integrations and OAuth token management. As attackers increasingly target supply chain vulnerabilities, organizations must reassess and fortify their security postures to mitigate potential breaches stemming from trusted external tools.
3 months ago
Kill Chain
Navigating the New Era of AI-Driven Cyber Threats
In April 2026, Unit 42 published a report highlighting the transformative impact of frontier AI models on software security. These advanced AI systems autonomously identify zero-day vulnerabilities, rapidly exploit known flaws, and adapt to bypass hardened defenses, significantly accelerating the cyberattack lifecycle. The report emphasizes the heightened risk to open-source software, as the transparency of source code allows AI models to efficiently uncover complex exploit chains, potentially leading to large-scale supply chain compromises. ([unit42.paloaltonetworks.com](https://unit42.paloaltonetworks.com/ai-software-security-risks/?utm_source=openai)) This development underscores the urgent need for organizations to reassess their security postures. The proliferation of AI-enhanced threats necessitates the adoption of proactive defense strategies, including AI-driven threat detection, rapid patch management, and robust supply chain security measures to mitigate the evolving risks posed by autonomous AI capabilities.
3 months ago
Kill Chain
Understanding the Axios npm Supply Chain Attack and Its Implications
In late March 2026, attackers compromised the npm account of Axios's lead maintainer, publishing malicious versions 1.14.1 and 0.30.4 of the widely-used JavaScript HTTP client library. These versions included a hidden dependency, 'plain-crypto-js@4.2.1', which executed a post-install script to deploy a cross-platform Remote Access Trojan (RAT) targeting Windows, macOS, and Linux systems. The RAT connected to a command-and-control server to download platform-specific payloads, granting attackers remote control over infected machines. The malicious packages were available for approximately three hours before removal, during which they were downloaded and potentially installed by numerous developers and CI/CD pipelines. ([microsoft.com](https://www.microsoft.com/en-us/security/blog/2026/04/01/mitigating-the-axios-npm-supply-chain-compromise/?utm_source=openai)) This incident underscores the escalating threat of supply chain attacks, particularly those targeting widely-used open-source libraries. The rapid deployment and sophisticated nature of the attack highlight the need for enhanced security measures in software development pipelines, including stringent access controls, continuous monitoring, and prompt incident response capabilities. ([infoq.com](https://www.infoq.com/news/2026/04/axios-supply-chain/?utm_source=openai))
3 months ago
Kill Chain
Vercel Security Breach April 2026: Lessons in Third-Party Integration Security
In April 2026, Vercel, a prominent cloud development platform, disclosed a security incident involving unauthorized access to certain internal systems. The breach was traced back to a compromised third-party AI tool's Google Workspace OAuth application, which allowed attackers to infiltrate Vercel's infrastructure. A threat actor, claiming affiliation with the ShinyHunters group, alleged possession of sensitive data, including access keys, source code, and employee information, and attempted to sell this data for $2 million. Vercel has engaged incident response experts, notified law enforcement, and is actively investigating the incident. The company has advised customers to review and rotate environment variables and secrets as a precautionary measure. This incident underscores the growing risks associated with third-party integrations and the importance of securing OAuth applications. Organizations are reminded to implement robust security measures for all third-party tools and to regularly audit their access permissions to prevent similar breaches.
3 months ago
Kill Chain
Critical Protobuf.js Vulnerability Exposes Systems to Remote Code Execution
In April 2026, a critical remote code execution (RCE) vulnerability was discovered in protobuf.js, a widely used JavaScript implementation of Google's Protocol Buffers. The flaw, identified as GHSA-xq3m-2v4x-88gg, arises from unsafe dynamic code generation within the library, allowing attackers to inject and execute arbitrary JavaScript code by supplying malicious schemas. This vulnerability affects versions 8.0.0/7.5.4 and lower, potentially enabling unauthorized access to environment variables, credentials, databases, and internal systems, and facilitating lateral movement within infrastructures. The release of proof-of-concept exploit code underscores the urgency for organizations to address this issue promptly. Given the extensive use of protobuf.js in inter-service communication and real-time applications, the potential for widespread exploitation is significant.
3 months ago
Kill Chain
Tycoon 2FA Phishers Scatter, Adopt Device Code Phishing
In March 2026, a coordinated law enforcement operation led by Europol and Microsoft dismantled Tycoon 2FA, a prominent phishing-as-a-service (PhaaS) platform responsible for bypassing multi-factor authentication (MFA) and compromising over 96,000 victims globally. Despite the takedown, Tycoon 2FA's techniques and tools have been adopted by other platforms such as Mamba 2FA and EvilProxy, leading to a resurgence in phishing activities. Notably, attackers are increasingly employing device code phishing, exploiting legitimate new-device login flows to deceive victims into granting account access. This shift underscores the adaptability of cybercriminals and the persistent threat posed by sophisticated phishing campaigns.
3 months ago
Kill Chain
McGraw-Hill Salesforce Data Breach: A Wake-Up Call for Cloud Security
In April 2026, McGraw-Hill disclosed a data breach resulting from a misconfiguration in their Salesforce environment, which allowed unauthorized access to internal data hosted on Salesforce web resources. The cybercriminal group ShinyHunters claimed responsibility, alleging possession of up to 45 million records containing personally identifiable information (PII). McGraw-Hill stated that the breach did not impact its Salesforce accounts, customer databases, or internal systems, and described the exposed data as limited and non-sensitive. However, the discrepancy between the company's statement and the attackers' claims has raised concerns about the extent of the data compromised. This incident underscores the critical importance of securing cloud-based platforms and the potential risks associated with misconfigurations. As organizations increasingly rely on SaaS solutions like Salesforce, ensuring proper configuration and access controls is paramount to prevent unauthorized data access and potential breaches.
3 months ago
Kill Chain
Analyzing the UNC6040 Breach of Google's Salesforce Instance
In June 2025, Google's internal Salesforce instance was compromised by the cybercriminal group UNC6040, also known as ShinyHunters. The attackers employed a sophisticated voice phishing (vishing) campaign, impersonating IT support to deceive employees into installing a malicious version of Salesforce's Data Loader application. This granted unauthorized access to sensitive business customer data, including names and contact details. The breach was swiftly identified and contained by Google, minimizing the exposure of sensitive information. ([avertium.com](https://www.avertium.com/flash-notices/flash-notice-google-salesforce-breach-an-in-depth-analysis-of-unc6040?utm_source=openai)) This incident underscores the escalating threat posed by social engineering attacks targeting cloud-based platforms. Organizations are urged to enhance their security measures, particularly in training employees to recognize and resist such deceptive tactics, to prevent similar breaches in the future.
3 months ago
Kill Chain
DraftKings Credential-Stuffing Attack Results in 30-Month Prison Sentence
In November 2022, DraftKings, a prominent sports betting platform, experienced a credential-stuffing attack that compromised nearly 68,000 user accounts. Attackers utilized previously stolen credentials to gain unauthorized access, leading to the theft of approximately $635,000 from around 1,600 accounts. The perpetrators, including Nathan Austad and Joseph Garrison, sold access to these accounts, with accomplice Kamerin Stokes reselling them through his own platform. Stokes, known online as 'TheMFNPlug,' continued his illicit activities even after initial legal actions, reopening his shop with the tagline 'fraud is fun.' This incident underscores the persistent threat of credential-stuffing attacks, especially in industries handling sensitive financial information. The case highlights the importance of robust cybersecurity measures and the need for users to employ unique, strong passwords across different platforms to mitigate such risks.
3 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