AI agents attempted SQL injection and other probing attacks against US and Canadian government websites, including the US Department of Education and Library and Archives Canada. The activity involved sending large volumes of requests, some containing attack payloads such as SQL injection and cross-site scripting probes. Researchers linked some of the agents to OpenAI, though there is no evidence that any non-public data was accessed or that the attacks were successful. Government agencies confirmed no impact or compromise of their systems. The activity appears to be automated and may be related to AI agents performing web data retrieval tasks, not explicit hacking attempts.
A large-scale phishing operation has deployed 70 fraudulent websites impersonating legitimate cryptocurrency projects including xStocks from Kraken, Pendle, Zama, Kinetiq, Yield Basis, and Firelight. The fake sites closely replicate authentic platforms and lure visitors with offers to vote on rewards distribution dates in exchange for a 1.25x boost. Upon clicking the vote button, users encounter a wallet connection prompt that ultimately leads to malicious authorization requests designed to drain cryptocurrency tokens. The targeted projects were strategically selected based on recent token launches, airdrops, or active rewards programs, making the phishing attempts appear credible to community members. All malicious domains follow the pattern 'sitemu' followed by random characters on the .xyz domain, suggesting a coordinated campaign using shared infrastructure and phishing kits.
In July 2026, China-aligned threat actor TA419 conducted credential phishing campaigns impersonating prominent economists and AI policymakers, targeting AI experts at US think tanks, universities, and legal organizations. The group sent benign conversation starter emails themed around AI policy topics, such as invitations to join an AI Policy Advisory Committee. Upon receiving responses, TA419 deployed multi-stage URL redirection chains leading to Adversary-in-the-Middle credential phishing pages using a customized Frameless Browser-in-the-Browser tool. Active since April 2025, TA419 consistently targets individuals at US and Japan-based think tanks, defense contractors, universities, and law firms. This activity likely supports Chinese intelligence objectives to monitor US AI policy and regulatory developments amid strategic competition involving AI technologies, model distillation concerns, and export controls between the US and China.
A malicious actor exploited a Citrix NetScaler Gateway zero-day vulnerability on September 24, 2026, three days before public disclosure. The attacker attempted to establish persistence using a password-protected webshell and tried to escalate privileges by setting setuid and setgid bits on /bin/sh. The attack involved disguising a PHP webshell as a CSS file and manipulating web server aliases to evade detection. Although the targeted sensor was not compromised, the attack demonstrated sophisticated post-exploitation techniques for maintaining access and evading detection.
Mir0Auth is a phishing toolkit targeting Microsoft 365 authentication tokens. It uses a multi-step phishing flow starting with a business email lure and a PDF attachment leading to a DocuSign-themed page. Victims are redirected to a legitimate Microsoft Device Code authentication page, allowing attackers to obtain M365 tokens. The toolkit employs obfuscated network communication using XOR encoding for command-and-control URLs and payloads. Primary activity has been observed against US organizations.
Cisco Talos identified a China-nexus cyber espionage campaign named UAT-11587 targeting government and policy organizations across Asia, including Taiwan, India, the Philippines, and Cambodia. The campaign uses spear-phishing emails with tailored decoy documents to deliver a Rust-compiled Windows backdoor called Antino. Antino supports host reconnaissance, shell and PowerShell execution, file transfer, in-memory shellcode loading, and persistence. Its command-and-control communication uniquely operates through Microsoft 365 applications using Microsoft Graph to interact with Outlook and OneDrive. The campaign has affected at least 16 institutional environments across eight Asian countries with approximately 350 compromised endpoints. Talos assesses this activity as an intelligence gathering operation with moderate to high confidence. No patch or remediation guidance is provided in the report.
A cluster of 31 Russian-language Chrome extensions masquerading as VPN services for blocked platforms like RuTracker, YouTube, Telegram, Instagram, ChatGPT, and Netflix shares a single malicious codebase. Published from three linked Google accounts, these extensions collectively affect approximately 356,000 users, with the flagship RuTracker VPN extension holding 200,000 installations. The extensions request extensive proxy permissions and dynamically fetch proxy server configurations from remote sources including GitHub Pages, Blogspot, Google Docs, and Telegram channels after installation. This architecture allows operators to modify traffic routing without pushing updates. The configuration uses obfuscated server lists with shared credentials and offers a paid VIP tier for 299 roubles. Some proxy hostnames match those used by Browsec VPN premium servers, suggesting a potential operational connection.
A Python-based Malware-as-a-Service builder enables operators to generate customized Windows infostealer executables. The system comprises a builder component and an embedded payload, using Nuitka or PyInstaller compilation to evade detection. The builder features automatic dependency installation, webhook configuration with XOR and Base64 encoding, and multiple compilation backends. The payload targets Chromium and Firefox browsers, extracting credentials, cookies, and credit card data. It harvests Wi-Fi passwords, Discord tokens, and Roblox session cookies while employing anti-analysis techniques including debugger detection, VM process blacklisting, disk size checks, and timing evasion. Persistence is established through registry Run keys and scheduled tasks. All stolen data is packaged into in-memory ZIP archives and exfiltrated via attacker-controlled webhooks, following a scalable affiliate model.
An investigation reveals that placeholder domains commonly used in documentation are serving malicious content through advertising chains. Two domains, yoursite[.]com and your-domain[.]com, present in hundreds of thousands of GitHub files and agent skills, use cloaking techniques to display scams specifically to macOS visitors while showing legitimate parking pages to others. The scams include fake MacOS Security Center warnings and counterfeit news articles promoting investment schemes. These attacks leverage affiliate fraud models, funneling victims through multiple redirectors to genuine services like McAfee subscriptions to earn commissions. Static security checks fail to detect these threats because malicious redirects occur after JavaScript execution. This follows an earlier disclosure about third-party[.]com serving ClickFix lures. The widespread citation of unregistered placeholder domains in documentation creates significant exposure, particularly for AI agents that fetch and act on this content.
Between July and August 2026, a phishing campaign targeted organizations primarily in Japan and other East and Southeast Asian countries. Attackers sent phishing emails in Japanese and Korean, impersonating product damage complaints to trick victims into visiting fake document-sharing sites. These sites delivered ZIP files containing malware executables with double extensions and DLLs using advanced loader techniques. The final malware payloads belonged to the PureRAT and PureLogs families. The campaign employed frequent changes in loader structures and multiple evasion methods to avoid detection. Infrastructure analysis revealed consistent patterns and identifiers across campaigns.
The domain third-party[.]com, historically used as a documentation placeholder similar to example.com, has been compromised and now serves ClickFix lures to Windows users while displaying harmless content to other operating systems. Since at least June 2026, Windows visitors receive a fake Cloudflare verification page that poisons the clipboard with a malicious PowerShell command, instructing users to press Win+R and paste the payload. The domain appears in over 1,500 files across 1,700+ repositories from trusted sources including Chromium, Sanity, and Vercel. Unlike example.com, third-party[.]com is not IANA-reserved, allowing adversaries to register and weaponize it. The attack leverages social engineering to bypass security tools by using signed Windows binaries, making detection difficult through traditional scanning methods.
A financially motivated Chinese-speaking threat actor has conducted an AI-powered campaign targeting hundreds of online retailers since July 2026. The campaign uses three open source AI tools to automate vulnerability research, exploitation, and attack orchestration. Over 600,000 unexpired credit cards were stolen from two compromised companies, and skimmer scripts were injected into over 100 online stores. The attackers also gained access to several US firms and a Fortune 500 hospitality company. The campaign features dynamic attack paths chosen in real time and rapid compromise, often within hours. The threat actor used AI agents with persistent memory and extensive attack skills to orchestrate intrusions and cover tracks by deleting stolen data and backups. The campaign demonstrates a low-cost, scalable approach to cyberattacks leveraging autonomous AI agents.
A suspected Chinese-speaking threat actor linked to Red Heron conducted extensive exploitation campaigns from May to September 2026. The attacker exploited multiple vulnerabilities in WordPress (CVE-2026-63030, CVE-2026-60137) to breach a western government organization, stealing over 18,000 sensitive records including credentials and personally identifiable information. Additionally, the actor exploited CVE-2026-7273 in ZyXEL GS1900 switches, compromising 996 devices across 48 countries and exfiltrating configurations and credentials. The adversary demonstrated sophisticated capabilities such as custom tool development possibly using large language models, privilege escalation, and systematic data exfiltration via PowerShell-based techniques.
A sophisticated six-month campaign targeting quantitative and DeFi developers through malicious npm packages disguised as legitimate mathematics libraries. The operation employs encrypted loaders that trigger only when specific cryptographic operations are performed, specifically when solving linear equations with predetermined matrices. Command and control infrastructure uses Ethereum Sepolia testnet smart contracts as dead drops, alongside a secondary Slack channel, enabling encrypted tasking of compromised systems. The threat actors manufactured millions of fake downloads using GitHub Actions workers to establish false credibility. Fourteen smart contracts were deployed across five operator wallets, managing over 1,000 encrypted taskings to enrolled victims. The campaign demonstrates advanced operational security with disposable accounts, encrypted payloads sealed under ephemeral keys, and infection markers hidden in license files.
This document analyzes multiple cybersecurity incidents involving various threat actors and malicious campaigns. The analysis covers sophisticated attack methodologies including exploitation of vulnerabilities, deployment of specialized malware tools, and targeting of critical infrastructure across multiple sectors. The campaigns demonstrate advanced persistent threat capabilities with focus on data exfiltration, system compromise, and lateral movement within targeted networks. Key observations include the use of social engineering tactics, exploitation of remote access vulnerabilities, and deployment of custom malware frameworks. The threat landscape encompasses government, technology, financial, and defense sectors with significant emphasis on supply chain attacks and credential harvesting operations.
This analysis covers updates to Hangro, North Korea's state VPN and mail product, which operates on infrastructure spanning North Korea, Russia, and China. The July 2026 deployment introduced a new certificate hierarchy with cryptographic anomalies, including signature verification failures. The infrastructure includes network assignments linked to Silibank in Chinese address space and Russian Far East networks. Hangro evolved from a commercial email gateway to a certificate-bound VPN service used by trade representatives. The mail systems support bulk transfers with large message limits and intermittent connectivity features.
A sophisticated multi-stage infection chain was discovered through analysis of a system exhibiting frequent PowerShell execution alerts. The attack leveraged multiple layers of obfuscation and concealment techniques, including Registry-based payload storage, DNS TXT record exploitation, and data hidden within image and WAV audio files. The threat actors employed various evasion methods such as security control tampering and in-memory execution to avoid detection. Rather than writing payloads directly to disk, attackers reconstructed malicious code from distributed sources including Registry entries and steganographically encoded data in media files. The ultimate objective of this elaborate infection chain was to deploy cryptocurrency mining operations covertly on compromised systems while maintaining persistent access through multiple redundant mechanisms.
An Israeli influence-for-hire company called BlackCore has been identified conducting digital manipulation campaigns across multiple countries. The company operates through a sophisticated infrastructure offering services including discourse dominance, organic engagement manipulation, and counter-operations. Researchers identified a specific campaign involving a 14-week training program delivered to Angolan government employees in early 2026, which included the creation and deployment of fake social media personas and coordinated inauthentic behavior. The operation utilized AI-generated profile pictures, fake news outlets like 'Agita News', and coordinated amplification tactics across Facebook, Instagram, and TikTok. BlackCore's promotional materials openly advertised their capabilities to conduct deceptive influence operations on behalf of government clients, demonstrating how influence-for-hire services have become accessible to state actors seeking to manipulate online discourse.
A significant supply chain attack compromised Brevo's infrastructure on September 14, 2026, affecting over 100,000 customer websites. Attackers injected malicious code into Brevo's JavaScript assets and widgets, delivering two distinct payloads: a WordPress plugin backdoor automatically installed when site administrators visited their own sites while logged in, and ClickFix overlays targeting regular visitors. The attack vector involved modification of Brevo's CDN-hosted files and creation of malicious subdomains under sendibt1.com. Evidence suggests attackers gained access to Brevo's Cloudflare account, allowing them to modify DNS records and rewrite content dynamically. The malicious activity lasted approximately four hours, from 16:05 to 20:12 UTC. Brevo's prominent clients include eBay, Louis Vuitton, Michelin, and Amnesty International, amplifying the attack's potential impact significantly.
