✨ The Containment Era is here. Secure AI workloads before they breach. →The Containment Era is here. →The Containment Era is here. →Explore ✨
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 Vulnerabilities Discovered in OFFIS DCMTK Toolkit Used in Medical Imaging
In June 2026, multiple critical vulnerabilities were identified in the OFFIS DCMTK Toolkit, a widely used DICOM toolkit in medical imaging software. These vulnerabilities, including CVE-2026-50003, CVE-2026-50254, CVE-2026-35505, CVE-2026-52868, and CVE-2026-44628, could allow attackers to write files outside intended directories, access unauthorized information, exhaust memory, or crash affected DCMTK client or server processes. The vulnerabilities affect DCMTK versions up to 3.7.0. ([hipaajournal.com](https://www.hipaajournal.com/offis-dcmtk-vulnerabilities-june-2026/?utm_source=openai)) The healthcare sector's reliance on DICOM standards for medical imaging makes these vulnerabilities particularly concerning. Exploitation could lead to unauthorized access to sensitive patient data and disruption of critical medical services. Organizations using affected versions are urged to apply the provided patches promptly to mitigate potential risks.
3 weeks ago
Kill Chain
MetaMask Users Targeted in Sophisticated Phishing Attack - July 2026
In July 2026, a sophisticated phishing campaign targeted MetaMask users by sending emails that falsely claimed their cryptocurrency wallets were at risk. The emails pressured recipients to provide their secret recovery phrases under the guise of securing their accounts. The attackers utilized a recently registered domain, captchasolve[.]help, to host the phishing site, effectively deceiving users into compromising their wallets. This incident underscores the evolving tactics of cybercriminals in exploiting user trust and the critical importance of safeguarding recovery phrases. ([isc.sans.edu](https://isc.sans.edu/diary/TA551%2B?utm_source=openai)) The prevalence of such targeted phishing attacks highlights the urgent need for enhanced user education on recognizing and avoiding social engineering schemes. As cryptocurrency adoption grows, both individuals and organizations must implement robust security measures and remain vigilant against deceptive practices that aim to exploit human vulnerabilities.
3 weeks ago
Kill Chain
Beware: Malicious 'Perplexity AI' Chrome Extension Intercepts User Searches
In June 2026, a malicious Chrome extension named "Search for perplexity ai" was discovered impersonating the legitimate Perplexity AI search engine. This extension altered users' default search settings, intercepting all address-bar queries and routing them through attacker-controlled infrastructure before redirecting to legitimate search services. While no credential theft was confirmed, the extension's permissions allowed for extensive data collection, posing significant privacy risks. ([bleepingcomputer.com](https://www.bleepingcomputer.com/news/security/fake-perplexity-extension-on-chrome-web-store-tracked-searches/amp/?utm_source=openai)) This incident underscores the growing trend of cybercriminals exploiting trusted AI brands to distribute malicious software. It highlights the need for enhanced vigilance in verifying browser extensions and the importance of robust security measures to prevent unauthorized data interception.
3 weeks ago
Kill Chain
Malicious PyPI Packages Compromise Telegram Bot Servers in 2026
Between November 2025 and June 2026, a campaign dubbed 'Operation Navy Ghost' targeted Python developers creating Telegram bots by distributing trojanized versions of the Pyrogram library on the Python Package Index (PyPI). These malicious packages, including 'VLifeGram' and 'pyrogram-styled', contained a hidden backdoor that, upon activation, allowed attackers to execute arbitrary code and access sensitive data on compromised servers. The backdoor was designed to operate silently, suppressing errors and disabling logging, thereby granting attackers extensive control over the affected systems. ([bleepingcomputer.com](https://www.bleepingcomputer.com/news/security/malicious-pypi-packages-give-hackers-control-of-telegram-bot-servers/?utm_source=openai)) This incident underscores the persistent threat of supply chain attacks in open-source ecosystems. The exploitation of widely-used libraries like Pyrogram highlights the need for developers to exercise caution when integrating third-party packages. Ensuring the integrity of software dependencies is crucial to prevent unauthorized access and data breaches.
3 weeks ago
Kill Chain
BioShocking Attack: A New Threat to AI Browser Security
In June 2026, researchers at LayerX identified a novel prompt injection attack named 'BioShocking' targeting AI-powered browsers. The attack involves a malicious webpage presenting a BioShock-themed puzzle game that rewards incorrect answers, conditioning the browser's control agent to disregard standard safety protocols. In the final stage, the agent is directed to access a GitHub repository and extract sensitive data, such as passwords. This proof-of-concept was tested against six mainstream agentic browsers, with only OpenAI's ChatGPT Atlas implementing an effective fix after disclosure. The BioShocking attack underscores the critical need for robust security measures in AI-driven applications. As AI agents become more integrated into daily tasks, their susceptibility to manipulation poses significant risks. This incident highlights the urgency for developers to implement explicit user confirmations for sensitive actions, enhance context checks, and establish strict boundaries for agentic sessions to prevent similar exploits.
3 weeks ago
Kill Chain
Langflow RCE Exploited to Deploy Monero Miner on Exposed AI App Endpoints
In March 2026, threat actors exploited a critical vulnerability in Langflow (CVE-2026-33017), an open-source AI workflow tool, to deploy Monero cryptocurrency miners on exposed AI application endpoints. This unauthenticated remote code execution flaw allowed attackers to execute arbitrary Python code via the public flow build API endpoint, leading to unauthorized system access and resource hijacking. The attacks were observed between March 27 and April 15, 2026, with malicious scripts terminating competing miners, disabling security controls, and establishing persistence mechanisms. ([thehackernews.com](https://thehackernews.com/2026/06/langflow-rce-exploited-to-deploy-monero.html?utm_source=openai)) The rapid exploitation of this vulnerability underscores the increasing targeting of AI infrastructure by cybercriminals. Organizations utilizing Langflow versions prior to 1.9.0 are urged to upgrade immediately and review their systems for signs of compromise. ([thehackernews.com](https://thehackernews.com/2026/06/langflow-rce-exploited-to-deploy-monero.html?utm_source=openai))
3 weeks ago
Kill Chain
Microsoft Identifies Critical AI Agent Vulnerability in MCP Tool Descriptions
In June 2026, Microsoft researchers identified a critical vulnerability in AI agents utilizing the Model Context Protocol (MCP). Attackers can exploit this by embedding malicious instructions within tool descriptions, causing AI agents to inadvertently exfiltrate sensitive company data without triggering security alerts. This method leverages the trust AI agents place in tool descriptions, leading to unauthorized data disclosures. This incident underscores the evolving threat landscape as AI agents become more integrated into business operations. Organizations must reassess their AI security protocols to address these sophisticated attack vectors, emphasizing the need for stringent validation of third-party tools and continuous monitoring of AI agent activities.
3 weeks ago
Kill Chain
Djinn Stealer Exploits SimpleHelp Vulnerability CVE-2026-48558
In June 2026, attackers exploited CVE-2026-48558, a critical authentication bypass vulnerability in SimpleHelp's OpenID Connect (OIDC) authentication flow, to gain unauthorized access to remote monitoring and management (RMM) systems. By submitting forged identity tokens, they obtained technician-level access without valid credentials, enabling them to deploy the Djinn Stealer malware across Windows, macOS, and Linux systems. Djinn Stealer targeted a wide range of sensitive information, including cloud service credentials, source control data, package registry credentials, AI development tools, and cryptocurrency wallets, posing significant risks to enterprise environments. This incident underscores the increasing focus of threat actors on exploiting vulnerabilities in trusted administrative tools to gain broad access to enterprise networks. The rapid exploitation of CVE-2026-48558 highlights the urgency for organizations to promptly apply security patches and implement robust monitoring to detect unauthorized access, especially in systems that manage critical infrastructure and sensitive data.
4 weeks ago
Kill Chain
Apple's June 2026 Security Updates: Addressing AI-Discovered WebKit Vulnerabilities
In June 2026, Apple released security updates for iOS, macOS, and Safari to address over 30 vulnerabilities, including four critical WebKit flaws discovered using AI tools like Anthropic Claude and OpenAI Codex. These vulnerabilities, such as CVE-2026-43707, involved memory corruption issues that could lead to unexpected crashes when processing malicious web content. Apple promptly addressed these issues with improved memory handling. ([thehackernews.com](https://thehackernews.com/2026/06/apple-patches-30-ios-macos-safari-flaws.html?utm_source=openai)) This incident underscores the growing role of AI in both identifying and potentially exploiting software vulnerabilities, highlighting the need for accelerated patch management and proactive security measures to mitigate emerging threats.
4 weeks ago
Kill Chain
BioShocking Attack: A Wake-Up Call for AI Browser Security
In June 2026, security firm LayerX unveiled a novel attack technique named 'BioShocking,' which exploits AI-powered browsers and assistants to extract user credentials. By presenting a malicious webpage designed as a puzzle game, attackers manipulated AI agents into disregarding their safety protocols. The agents, including OpenAI's ChatGPT Atlas, Perplexity's Comet, and Anthropic's Claude browser extension, were deceived into accessing and transmitting sensitive information, such as SSH login credentials from users' GitHub repositories, to unauthorized parties. This method leverages indirect prompt injection, where malicious commands are embedded within seemingly benign content, leading AI agents to execute unauthorized actions without detection. ([thehackernews.com](https://thehackernews.com/2026/06/new-bioshocking-attack-tricks-ai.html?utm_source=openai)) The BioShocking attack underscores the pressing need for enhanced security measures in AI-driven applications. As AI browsers become more integrated into daily workflows, their potential to access and manipulate sensitive data grows, making them attractive targets for cybercriminals. This incident highlights the importance of implementing robust safeguards, such as explicit user confirmations before accessing logged-in accounts and setting strict boundaries on AI agent capabilities, to prevent similar exploits in the future. ([layerxsecurity.com](https://layerxsecurity.com/blog/bioshocking-ai-gaming-the-ai-browser-and-escaping-its-guardrails/?utm_source=openai))
4 weeks ago
Kill Chain
Critical SimpleHelp Vulnerability (CVE-2026-48558) Exposes Systems to Unauthorized Access
In June 2026, a critical authentication bypass vulnerability, CVE-2026-48558, was identified in SimpleHelp versions 5.5.15 and earlier, as well as 6.0 pre-release versions. This flaw resides in the OpenID Connect (OIDC) authentication flow, where identity tokens are accepted without verifying their cryptographic signatures. Exploiting this vulnerability, remote unauthenticated attackers can forge tokens to gain full technician access, potentially bypassing multi-factor authentication and compromising managed endpoints. ([helpnetsecurity.com](https://www.helpnetsecurity.com/2026/06/16/simplehelp-rmm-cve-2026-48558/?utm_source=openai)) The exploitation of CVE-2026-48558 underscores the escalating risks associated with remote monitoring and management (RMM) tools, which are increasingly targeted by cybercriminals. Organizations must prioritize the timely application of security patches and enhance monitoring of authentication mechanisms to mitigate such threats.
4 weeks ago
Kill Chain
Study Reveals 282 iOS AI Apps Exposing LLM API Credentials
In June 2026, researchers from Wake Forest University analyzed 444 AI-powered iOS applications and discovered that 282 of them (approximately 64%) exposed exploitable LLM API credentials through network traffic. The vulnerabilities were categorized into three main types: plaintext API key transmission (19%), unauthenticated backend proxy access (33%), and JWT-based token leakage (48%). These security lapses allowed unauthorized access to AI services, potentially leading to financial losses for developers and compromising user data. Notably, only 28% of the affected applications had addressed these issues three months after being notified. This incident underscores a systemic gap in secure integration practices within the iOS ecosystem, highlighting the urgent need for developers to implement robust security measures when integrating large language models into mobile applications. The prevalence of such vulnerabilities calls for enhanced developer awareness, explicit security guidance from AI service providers, and potential platform-level enforcement to mitigate risks associated with credential leakage.
4 weeks 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