✨ 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
OpenSSF Tech Talk Recap: Securing Agentic AI
On April 8, 2026, the Open Source Security Foundation (OpenSSF) hosted a Tech Talk titled 'Securing Agentic AI,' addressing the unique security challenges posed by non-deterministic AI agents. Experts from Microsoft, Thread AI, Canonical, and the OpenSSF AI/ML Security Working Group discussed issues such as agent autonomy, tool-model trust, and context integrity. They introduced SAFE-MCP, a threat catalog inspired by the MITRE ATT&CK framework, detailing over 80 attack techniques targeting tool-based Large Language Models (LLMs). The session also emphasized the importance of securing the entire AI infrastructure stack, from user interfaces to hardware, highlighting the critical role of open source in each layer. ([openssf.org](https://openssf.org/blog/2026/04/08/openssf-tech-talk-recap-securing-agentic-ai/?utm_source=openai)) The relevance of this discussion is underscored by recent developments in AI security. For instance, Anthropic's AI model, Claude Mythos, identified thousands of zero-day vulnerabilities across major operating systems and web browsers, some unpatched for decades. This highlights the pressing need for robust security measures in AI systems to prevent potential exploitation. ([tomshardware.com](https://www.tomshardware.com/tech-industry/artificial-intelligence/anthropics-latest-ai-model-identifies-thousands-of-zero-day-vulnerabilities-in-every-major-operating-system-and-every-major-web-browser-claude-mythos-preview-sparks-race-to-fix-critical-bugs-some-unpatched-for-decades?utm_source=openai))
3 months ago
Kill Chain
Anthropic's Claude Mythos Preview: A Game-Changer in Cybersecurity
In April 2026, Anthropic unveiled its advanced AI model, Claude Mythos Preview, capable of autonomously identifying and exploiting thousands of zero-day vulnerabilities across major operating systems and web browsers. This unprecedented capability led to the launch of Project Glasswing, a collaborative initiative with tech giants like Amazon, Apple, and Microsoft, aiming to secure critical software infrastructure. Due to the model's potential for misuse, Anthropic restricted its access to select organizations, emphasizing the need for responsible deployment of such powerful AI tools. ([anthropic.com](https://www.anthropic.com/project/glasswing?utm_source=openai)) The emergence of AI models like Claude Mythos Preview signifies a paradigm shift in cybersecurity, where AI can both uncover and exploit vulnerabilities at an unprecedented scale. This development underscores the urgency for the cybersecurity industry to adapt, emphasizing proactive defense strategies and collaborative efforts to mitigate potential threats posed by advanced AI capabilities. ([red.anthropic.com](https://red.anthropic.com/2026/mythos-preview/?utm_source=openai))
3 months ago
Kill Chain
Claude Mythos: A New Era of AI-Driven Cybersecurity Threats
In early 2026, Anthropic's AI model, Claude Mythos, demonstrated unprecedented capabilities in autonomously identifying and exploiting software vulnerabilities across major operating systems and web browsers. This advancement significantly reduces the time between vulnerability disclosure and exploitation, posing a substantial challenge to traditional cybersecurity defenses. ([tomshardware.com](https://www.tomshardware.com/tech-industry/artificial-intelligence/anthropics-latest-ai-model-identifies-thousands-of-zero-day-vulnerabilities-in-every-major-operating-system-and-every-major-web-browser-claude-mythos-preview-sparks-race-to-fix-critical-bugs-some-unpatched-for-decades?utm_source=openai)) The emergence of AI-driven tools like Claude Mythos signifies a paradigm shift in cyber threats, enabling attackers to automate complex exploits and outpace defenders. Organizations must urgently adapt their security strategies to address these rapidly evolving AI-enhanced threats. ([axios.com](https://www.axios.com/2026/04/10/anthropic-mythos-openai-cyber-threats?utm_source=openai))
3 months ago
Kill Chain
Storm Infostealer 2026: A New Era of Cyber Threats
In early 2026, a new infostealer malware named 'Storm' emerged, enabling attackers to bypass traditional security measures by exfiltrating encrypted browser data to remote servers for decryption. This method allows the malware to harvest sensitive information such as saved passwords, session cookies, and cryptocurrency wallets without triggering endpoint security alerts. Storm's capabilities extend to automating session hijacking, granting attackers authenticated access to various platforms without the need for passwords or multi-factor authentication. The malware is offered as a subscription service, with packages starting at $300 for a 7-day demo and up to $1,800 for a full team license supporting 100 operators. Notably, data exfiltration continues even after subscriptions expire. The emergence of such turnkey hacking tools underscores the growing accessibility of sophisticated cyberattacks, posing serious risks to organizations relying solely on basic endpoint protections. Advanced behavioral and network analytics are essential for detecting such threats.
3 months ago
Kill Chain
OpenAI's 2026 Supply Chain Attack: Lessons in Software Security
In March 2026, OpenAI's macOS code-signing workflow was compromised due to a supply chain attack involving the widely used JavaScript library, Axios. The attackers, identified as the North Korean threat group UNC1069, gained access to the Axios maintainer's account and published malicious versions of the package. These versions were inadvertently incorporated into OpenAI's GitHub Actions workflow, potentially exposing code-signing certificates used for macOS applications such as ChatGPT Desktop, Codex, Codex CLI, and Atlas. Although OpenAI's investigation found no evidence of certificate misuse or compromise of user data, the company proactively revoked and rotated the affected certificates to mitigate any potential risks. This incident underscores the escalating threat of supply chain attacks targeting widely used open-source libraries. Organizations must remain vigilant, as such attacks can infiltrate even well-secured development pipelines, leading to potential downstream compromises. The involvement of state-sponsored actors like UNC1069 highlights the need for enhanced security measures and continuous monitoring of software dependencies to protect against sophisticated cyber threats.
3 months ago
Kill Chain
FBI Dismantles W3LL Phishing Platform in 2026
In April 2026, the FBI's Atlanta Field Office, in collaboration with Indonesian authorities, dismantled the 'W3LL' phishing platform, a sophisticated cybercrime operation that enabled attackers to create convincing replicas of corporate login portals. This platform facilitated the theft of thousands of credentials and was linked to over $20 million in fraud attempts. The operation led to the seizure of critical infrastructure and the arrest of the alleged developer, marking a significant milestone in international cybercrime enforcement. The takedown of W3LL underscores the escalating threat posed by Phishing-as-a-Service platforms, which lower the barrier to entry for cybercriminals and amplify the scale of attacks. This incident highlights the urgent need for organizations to enhance their cybersecurity measures, particularly in defending against advanced phishing techniques that can bypass multi-factor authentication and compromise sensitive data.
3 months ago
Kill Chain
OpenAI's Response to the 2026 Axios Supply Chain Attack
In late March 2026, OpenAI identified a security incident involving a compromised version of the Axios library, a widely used third-party developer tool. On March 31, a GitHub Actions workflow utilized in OpenAI's macOS app-signing process downloaded and executed the malicious Axios version 1.14.1. This workflow had access to critical code-signing certificates used for authenticating OpenAI's macOS applications, including ChatGPT Desktop, Codex App, Codex CLI, and Atlas. Despite the potential risk, OpenAI's investigation concluded that there was no evidence of user data access, system compromise, or software alteration. As a precautionary measure, OpenAI revoked and rotated the affected certificates and required all macOS users to update their applications to the latest versions by May 8, 2026, after which older versions would no longer receive support or function properly. This incident underscores the growing threat of supply chain attacks targeting widely used open-source libraries and developer tools. The compromise of a single library can have cascading effects across numerous organizations, highlighting the need for stringent security practices in software development pipelines. Organizations are urged to implement measures such as pinning dependencies to specific versions, conducting regular security audits, and maintaining robust incident response plans to mitigate such risks.
3 months ago
Kill Chain
APT37's Facebook Social Engineering Tactics Unveiled
In April 2026, the North Korean state-sponsored hacking group APT37 (also known as ScarCruft) initiated a sophisticated social engineering campaign targeting individuals via Facebook. The attackers created fake profiles to befriend targets, eventually moving conversations to Facebook Messenger. They persuaded victims to install a tampered version of Wondershare PDFelement, claiming it was necessary to view encrypted military documents. This malicious software executed embedded shellcode upon launch, establishing a foothold for the attackers. The campaign utilized compromised infrastructure for command-and-control operations, leveraging a legitimate Japanese real estate website to issue malicious commands. Ultimately, the malware was disguised as a harmless JPG image, enabling extensive remote access capabilities while evading detection by security software. This incident underscores the evolving tactics of APT37, highlighting their ability to exploit social media platforms for initial access and their use of legitimate software and infrastructure to evade detection. The campaign's success emphasizes the need for heightened awareness and robust security measures against social engineering attacks, especially those leveraging trusted platforms and applications.
3 months ago
Kill Chain
Anthropic's Claude Mythos AI Uncovers Thousands of Zero-Day Vulnerabilities
In April 2026, Anthropic unveiled its advanced AI model, Claude Mythos Preview, which autonomously identified thousands of zero-day vulnerabilities across major operating systems and web browsers. Notably, the model discovered a 27-year-old bug in OpenBSD and a 16-year-old flaw in FFmpeg's H.264 codec. Due to the potential risks associated with these findings, Anthropic restricted access to the model, collaborating with over 50 organizations, including tech giants like Amazon, Google, and Microsoft, under Project Glasswing to address and patch these vulnerabilities. This incident underscores the dual-edged nature of AI in cybersecurity, highlighting its potential to both uncover and exploit critical software flaws. The rapid advancements in AI capabilities necessitate a reevaluation of security protocols and the development of robust safeguards to prevent misuse. Organizations must stay vigilant and adapt to the evolving threat landscape shaped by AI-driven tools.
3 months ago
Kill Chain
SentinelOne's AI EDR Thwarts Zero-Day Supply Chain Attack Involving Anthropic's Claude AI
In March 2026, SentinelOne's AI-driven Endpoint Detection and Response (EDR) system autonomously identified and halted a zero-day supply chain attack involving a trojanized version of LiteLLM, a widely used proxy for LLM API calls. The compromised package, updated by Anthropic's Claude AI coding assistant without human intervention, attempted to execute malicious Python code across multiple customer environments. SentinelOne's Singularity Platform detected and blocked the payload before execution, preventing data theft, persistence, Kubernetes lateral movement, and encrypted exfiltration within hours of the attack's initiation. This incident underscores the escalating sophistication of supply chain attacks, particularly those exploiting AI-driven development tools. The rapid detection and mitigation by autonomous security systems highlight the necessity for organizations to adopt AI-native defenses capable of operating at machine speed to counteract evolving cyber threats.
3 months ago
Kill Chain
Identity-Based Attacks: The Predominant Cyber Threat in 2026
In 2026, identity-based attacks have emerged as the predominant cyber threat, with 67% of incidents involving compromised credentials, session tokens, or other forms of digital identity. Attackers increasingly exploit legitimate access methods, bypassing traditional security measures to infiltrate systems undetected. This shift underscores the critical need for organizations to enhance identity security protocols and adopt continuous monitoring strategies to detect and mitigate unauthorized access. The rise of identity-driven intrusions is further exacerbated by the integration of AI technologies, which enable adversaries to automate and scale their attacks more effectively. As a result, businesses must prioritize robust identity governance and implement advanced detection mechanisms to safeguard against these evolving threats.
3 months ago
Kill Chain
APT41's 2026 Cloud Credential Theft: A Wake-Up Call for Cloud Security
In early 2026, the Chinese state-sponsored threat group APT41 launched a sophisticated campaign targeting Linux-based cloud environments across AWS, Google Cloud Platform, Microsoft Azure, and Alibaba Cloud. Utilizing an undetectable ELF backdoor, the group harvested cloud credentials and metadata, enabling unauthorized access and potential data exfiltration. The malware employed typosquatting techniques and covert command-and-control channels over SMTP port 25, effectively evading traditional detection mechanisms. This incident underscores the evolving tactics of APT41, highlighting their focus on cloud infrastructure and the challenges in detecting such advanced persistent threats. Organizations must enhance their cloud security measures, monitor for anomalous activities, and implement robust detection strategies to mitigate similar threats.
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