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 AI Security Gap: DeepSeek Harness Sandbox Escape Exposes Autonomous Agent Risks
In August 2026, a critical vulnerability (CVE-2026-82533) was discovered in DeepSeek Harness, an open-source AI coding agent tool with over 216,000 GitHub stars. The flaw allowed sandboxed AI agents to disable their own security sandbox through a single command, bypassing file system protections designed to prevent untrusted code execution. Attackers could exploit this by supplying malicious text that prompted the agent to call the tool's local web interface, switching to 'danger-full-access' mode without approval prompts. The vulnerability stemmed from inadequate authentication on the local interface and improper host header validation, enabling both local sandbox escapes and potential remote exploitation through port forwarding. This incident highlights the growing security challenges in AI development tools as organizations increasingly adopt autonomous coding agents. The vulnerability demonstrates how AI agents can be manipulated to bypass their own safety mechanisms, representing a new class of security risks in the rapidly expanding AI development ecosystem.
1 week ago
Kill Chain
Critical Alby Hub Vulnerability Exposed Bitcoin Wallets to Complete Takeover
In September 2026, Bitcoin wallet company Alby disclosed a critical vulnerability in Alby Hub versions v1.7.0 through v1.18.5 that allowed attackers to completely take over internet-exposed Lightning wallets and drain funds. The flaw affected self-hosted Bitcoin wallets where owners had inadvertently exposed their Hub management interfaces to the public internet, often following Alby's own documentation that recommended such configurations. At least one user was confirmed affected, with the company providing limited details about the vulnerability mechanism pending responsible disclosure. The incident highlights the ongoing security challenges in cryptocurrency infrastructure, particularly as Bitcoin adoption accelerates and self-custody solutions become more mainstream. With ransomware groups increasingly targeting cryptocurrency platforms and the rise of state-sponsored attacks on financial infrastructure, vulnerabilities in wallet software present critical risks to both individual users and the broader digital asset ecosystem.
1 week ago
Kill Chain
Massive Infostealer Campaign Exposes Thousands of AI Service Tokens, Bypassing MFA Protection
In August 2026, cybersecurity researchers analyzed a 7GB infostealer dump containing data from 5,871 infected machines across 162 countries, revealing thousands of unexpired authentication tokens for AI services including Google, OpenAI, Anthropic, and others. Information stealers like Lumma Stealer and Vidar harvested session tokens, API keys, and JSON Web Tokens (JWTs) that threat actors can replay to bypass credential-based authentication and multi-factor authentication, effectively gaining unauthorized access to premium AI services without traditional login processes. The stolen data included 555 AI-related JWTs and 2,937 encrypted tokens, with 17.7% containing plaintext personally identifiable information, enabling account takeovers, resource theft, and unauthorized AI service usage sold on underground markets. This incident highlights the growing cybercriminal focus on AI credential theft as premium model access costs create strong financial incentives for stealing rather than purchasing legitimate access, while the proliferation of anti-detect browsers and session replay tools makes monetizing these stolen tokens increasingly accessible to threat actors.
1 week ago
Kill Chain
Critical Vulnerability in Lean Theorem Prover Enables Fabrication of Mathematical Proofs
Trail of Bits researchers discovered a critical vulnerability in Lean 4 theorem prover versions up to 4.33.1 that allowed fabrication of mathematical proofs through string manipulation exploits. The flaw in String.Pos.Raw.extract function created inconsistencies between logical definitions and compiled native code, enabling attackers to manufacture contradictions and prove false theorems, including a bogus proof of Fermat's Last Theorem. This supply-chain vulnerability affects the integrity of formal verification systems used in critical software development and mathematical research. This incident highlights the emerging risks in AI-assisted code generation and formal verification tools as they become integral to software supply chains. With increasing reliance on theorem provers for security-critical applications, vulnerabilities in these foundational tools pose systemic risks to mathematical proofs and software verification processes.
1 week ago
Kill Chain
Chinese AI Giants Caught in Massive U.S. Model Distillation Campaign
In 2024, Chinese AI companies including DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun, and Z.AI engaged in systematic distillation of U.S. frontier AI models according to a joint advisory from NSA, CISA, and FBI. The companies spent billions of tokens across millions of exchanges with models like Claude, ChatGPT, Google Gemini, and xAI's Grok to extract proprietary capabilities and strengthen their domestic AI systems. The attackers used sophisticated evasion techniques including distributed accounts, proxy networks, third-party aggregators, and gray market access to circumvent geographic restrictions and detection mechanisms. This incident highlights the growing threat of AI-enabled economic espionage as artificial intelligence becomes central to national competitiveness, with state-sponsored actors leveraging legitimate AI APIs for large-scale intellectual property theft through automated distillation campaigns.
1 week ago
Kill Chain
DoppelCart Fraud Network Exposes Massive E-Commerce Security Gap
In September 2026, cybersecurity researchers discovered DoppelCart, the largest documented fake e-commerce network comprising over 119,000 fraudulent domains, primarily using the .SHOP TLD. The operation impersonates 44,182 legitimate brands by copying product catalogs and branding, then uses WebSocket connections to steal payment card data, CVV codes, and personal information in real-time during checkout. The network affects 2.72% of all .SHOP domains and significantly surpasses previous operations like BogusBazaar's 75,000 sites, with over 105,000 shops remaining active at discovery. This incident highlights the evolving sophistication of financial fraud networks and their ability to operate at unprecedented scale through automated domain generation and brand impersonation, representing a critical threat to consumer trust and e-commerce security.
1 week ago
Kill Chain
StyleSmuggler Zero-Day: How CVE-2026-75650 Compromised Magento E-commerce Security
In September 2026, threat actors began exploiting CVE-2026-75650 (StyleSmuggler), a critical zero-day vulnerability in Adobe Commerce and Magento Open Source with a CVSS score of 10.0. The flaw allows unauthenticated remote code execution through PHP code injection in Magento's template system, specifically targeting the 'Payment Transaction Failed Reminder' email function. Attackers deployed sophisticated payloads including a Rust-based Linux backdoor and PHP web shells on compromised e-commerce sites, with exploitation beginning September 4, 2026. This incident highlights the growing sophistication of supply chain attacks targeting e-commerce platforms, as attackers increasingly focus on high-value web applications that process financial transactions and customer data, demonstrating the critical need for runtime application security controls.
1 week ago
Kill Chain
Grindr's £26 Million Settlement Exposes Critical Gaps in Dating App Data Privacy
In September 2026, Grindr agreed to pay £26 million ($35.1 million) to settle a U.K. class action lawsuit involving over 10,000 users whose sensitive personal data, including HIV status, was shared with third-party advertising companies Apptimize and Localytics between 2018-2020. The incident, originally exposed by Norwegian research group SINTEF in April 2018, occurred while Grindr was owned by Chinese gaming company Kunlun, before its 2020 acquisition by San Vicente Acquisition LLC. The settlement covers historical data practices that violated U.K. privacy laws through unauthorized sharing of location data, sexual orientation, and medical information for commercial advertising purposes. This incident highlights the ongoing regulatory scrutiny of data privacy violations in dating apps and social platforms, particularly as GDPR enforcement intensifies and class action lawsuits become more prevalent in addressing historical privacy breaches involving sensitive personal information.
1 week ago
Kill Chain
BengalSEO Campaign Weaponizes Search Results for MayaBot Distribution
The BengalSEO campaign represents a sophisticated search engine optimization poisoning operation that has been active since 2015, targeting Bing search results to deliver MayaBot malware and facilitate tech support scams. Operating from Rajasthan, India, the threat actors behind this campaign manipulate search engine results to redirect victims to malicious websites, where they deploy malware or engage in fraudulent technical support schemes. The campaign demonstrates the evolution of SEO poisoning techniques and their effectiveness in reaching unsuspecting users through legitimate search queries. This incident highlights the growing sophistication of search engine manipulation attacks and their integration with traditional malware distribution methods. As organizations increasingly rely on digital visibility and search engine optimization, the weaponization of these same techniques by threat actors represents a significant shift in attack vectors that security teams must address.
1 week ago
Kill Chain
When AI Attacks AI: The 2026 Autonomous Agent Credential Harvesting Campaign
In September 2026, threat actors deployed autonomous AI agent frameworks to conduct large-scale credential harvesting operations, compromising thousands of third-party credentials in under six hours. Google Threat Intelligence Group identified multiple financially motivated groups, including TeamPCP, leveraging AI-assisted tools like DUSTMAKER malware to target AI coding assistants, cloud environments, and supply chains across PyPI, npm, and Docker Hub repositories. The attacks demonstrated unprecedented automation capabilities, with AI systems autonomously managing vulnerability scanning, real-time troubleshooting, and IP rotation without human intervention. This incident represents a critical escalation in AI-enabled cyber threats, coinciding with the rapid adoption of generative AI tools in enterprise environments and the emergence of 'abliterated' open-weight models that bypass safety guardrails.
1 week ago
Kill Chain
ChatGPT Prompt Injection Flaw Exposed Gmail Data Through Hidden Cross-Account Channels
In September 2026, Check Point Research disclosed a critical vulnerability in OpenAI's ChatGPT that allowed attackers to inject malicious prompts that could silently exfiltrate user data from connected applications like Gmail. The attack exploited a shared internal JFrog Artifactory service used by ChatGPT's isolated containers, creating an unauthorized communication channel between different user accounts. Attackers could plant instructions through shared conversations, custom GPTs, or user-pasted prompts that would execute hidden data theft operations while displaying normal responses to victims. OpenAI confirmed the vulnerability and took the internal service offline after disclosure. This incident highlights the emerging risks of AI systems as attack vectors, particularly as organizations increasingly integrate AI tools with sensitive business applications and data sources, making prompt injection attacks a critical new threat category requiring immediate security attention.
1 week ago
Kill Chain
WeChat Zero-Click Worm: How 1.4 Billion Users Were at Risk from Incoming Calls
In July 2026, security researchers at Calif discovered a critical zero-click vulnerability in WeChat that allowed attackers to take complete control of user accounts through incoming calls without any user interaction. The exploit worked by leveraging WeChat's contact trust system, enabling worm-like propagation where compromised accounts could automatically infect other contacts. Affecting WeChat's 1.4 billion user base across iPhone and Android platforms, the vulnerability granted attackers full access to messages, payments, and WeChat's extensive ecosystem of mini-programs and services. Tencent patched the flaw in August 2026 versions 8.0.77 for Android and 8.0.76 for iOS. This incident highlights the growing sophistication of mobile application attacks and the critical importance of securing communication platforms that serve as digital wallets and business ecosystems, particularly as zero-click exploits become increasingly weaponized against high-value messaging applications.
1 week ago
Kill Chain
Stop Active Cloud Data Exfiltration
Aviatrix Rapid Containment 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