Intelligence Feed

1mo ago · hacker-news

ACR Stealer, an infostealer active since 2024, is being distributed through social engineering lures such as fake Claude AI assistant pages and malvertising. The malware uses fileless techniques and WebDAV shares to steal browser credentials, session tokens, and sensitive files from Microsoft 365, OneDrive, and SharePoint. It relies on user execution via pasted commands and does not exploit software vulnerabilities, making detection dependent on behavioral analysis and proactive controls.

1mo ago · hacker-news

Armenia has detained a Russian national, Aleksandr Ermakov, based on a U.S. extradition request related to alleged involvement in REvil (Sodinokibi) ransomware attacks. However, his legal team claims authorities have mistaken him for another individual with the same name—Aleksandr Gennadievich Ermakov—who was sanctioned in 2024 for cybercriminal activity including the Medibank data breach. The accused individual in custody, Aleksandr Yuryevich Ermakov, is a former prison-service lawyer with no known English proficiency, and lawyers argue insufficient identification was used for the arrest. The confusion stems from identical names and lack of patronymic or biometric verification in the extradition process.

1mo ago · hacker-news

The article discusses the growing emphasis on military autonomy and the need for trusted information infrastructure to support interconnected autonomous systems across defense forces. It highlights strategic investments by the U.S., UK, and NATO in autonomous technologies and underscores the importance of secure, cross-domain data sharing for mission effectiveness. The focus is shifting from deploying individual autonomous platforms to enabling secure, trusted, and interoperable information exchange across systems and coalition partners. The article promotes Everfox’s hardware-enforced separation solution as a means to achieve secure, high-assurance connectivity for autonomous military operations.

1mo ago · hacker-news

The European Commission has mandated Google to open access to Android's microphone, camera, screen, and other core features for rival AI assistants under the Digital Markets Act. This decision requires Google to implement a certification program allowing third-party AI assistants to access sensitive device functions, while also enforcing strict security and privacy safeguards. The move aims to promote competition but raises concerns about potential abuse of powerful device permissions by untrusted applications. Google must comply by August 2027, with certain features deferred to 2028.

1mo ago · hacker-news

The U.S. Cybersecurity and Infrastructure Security Agency (CISA) has added the actively exploited SharePoint Server remote code execution vulnerability CVE-2026-58644 to its Known Exploited Vulnerabilities (KEV) catalog. This critical zero-day flaw allows authenticated attackers with Site Owner privileges to execute arbitrary code remotely on vulnerable SharePoint servers. CISA warns of ongoing exploitation and mandates federal agencies to apply patches by July 19, 2026. Additional SharePoint-related vulnerabilities are also being actively exploited, enabling remote code execution and post-exploitation activities such as theft of IIS machine keys.

1mo ago · hacker-news

Owen Flowers and Thalha Jubair, alleged members of the Scattered Spider threat actor group, were sentenced to five and a half years each for their roles in a 2024 cyberattack on Transport for London (TfL) that disrupted 148 systems and affected 27,000 employees. The attack, which exfiltrated personal and financial data and nearly led to a network shutdown, caused £29 million in losses and recovery costs. The hackers also targeted US healthcare organizations, with threats to disrupt critical systems, and are linked to a broader campaign of social engineering, SIM swapping, and data extortion spanning hundreds of attacks from 2022 to 2025.

1mo ago · hacker-news

ClickLock is a new macOS infostealer that uses social engineering via a fake Cloudflare CAPTCHA to trick users into pasting a malicious command into Terminal. Upon refusal to enter credentials, it initiates aggressive app-killing loops every 210ms to coerce compliance. It steals login passwords, browser credentials, crypto wallets, and Keychain data, exfiltrating via Telegram bots. The malware uses compromised websites for payload delivery and a modified open-source backdoor, with persistence via LaunchAgents.

1mo ago · hacker-news

A vulnerability in n8n's Enterprise token exchange feature, tracked as CVE-2026-59208, allowed attackers to log in as users from another issuer due to improper validation of JWT tokens. The flaw occurred when n8n matched incoming tokens solely on the 'sub' claim without verifying the 'iss' (issuer), enabling account takeover if two trusted issuers used overlapping subject identifiers. The issue affects n8n versions prior to 2.27.4 and 2.28.1, and while the feature is limited to Enterprise deployments in preview, it poses a high-severity risk for misconfigured systems.

1mo ago · hacker-news

Multiple cyber threat campaigns were observed in mid-2026, including malicious NuGet packages distributing spyware disguised as game cheats, fake installers delivering the Starland RAT and WLDR C2 implant, and a new ransomware family named Spirals that encrypted a South Asian IT firm's network within 24 hours. Threat actors exploited known vulnerabilities such as CVE-2026-46817 and CVE-2023-4346, while also leveraging social engineering via phishing eCards and OAuth device code attacks. Additional threats include large-scale infostealer distribution through fake GitHub repositories, Chrome Sync abuse for stalking, and dual monetization campaigns deploying Vidar stealer and XMRig miner.

1mo ago · hacker-news

A critical unpatched vulnerability in Shark robot vacuums allows attackers with access to a device's certificate to execute arbitrary commands on other vacuums within the same AWS region. The flaw stems from overly permissive AWS IoT policies that permit wildcard subscription and publishing to device shadows, enabling remote code execution, camera access, and Wi-Fi password theft. The issue affects older Shark vacuum models whose certificates were issued with unrestricted policies, and the fix requires server-side policy updates from SharkNinja, as no firmware update is needed. Despite being reported in March 2026, no patch or CVE has been issued as of July.

1mo ago · hacker-news

The article discusses the growing reliance on AI in offensive security and highlights the risks of treating AI-generated findings as validated vulnerabilities without proper human verification. It emphasizes that while AI can accelerate vulnerability discovery, it cannot replace human judgment in proving exploitability, impact, and real-world risk. The core message is that validation through technical knowledge and reproducible evidence remains essential to distinguish noise from genuine threats.

1mo ago · hacker-news

A China-linked threat actor has resurfaced with the Daxin kernel-mode rootkit, detected in 2026 within a Taiwan-based subsidiary of a multinational high-tech manufacturer. The compromised system was also infected with a previously undocumented backdoor, Stupig, which enables pre-login SYSTEM-level command execution by masquerading as a legitimate keyboard DLL. Both malware samples were compiled in 2013, suggesting long-term stealthy persistence, with Daxin using covert C2 via hijacked TCP connections and Stupig enabling credential theft before user login. The attack highlights sophisticated, sustained cyber espionage activity targeting critical infrastructure.

1mo ago · hacker-news

Researchers have identified a new class of attack called Agent Data Injection (ADI), which exploits how AI agents parse structured data by injecting malicious input disguised as trusted data elements like sender names or button IDs. Unlike traditional prompt injection, ADI corrupts underlying facts the agent trusts, enabling actions such as unintended clicks or execution of attacker-controlled commands. The attack affects multiple AI models including GPT-5, Claude, and Gemini, with success rates up to 50% despite existing defenses. No real-world exploitation has been reported, but proof-of-concept demonstrations show high effectiveness across web and coding agents.

1mo ago · hacker-news

The PhantomEnigma campaign has hijacked over 20 Brazilian government websites, leveraging compromised .gov.br domains and authenticated email accounts to distribute malware. The attack uses fake police-themed documents to lure victims, redirecting them through trusted government infrastructure to deliver a modular Inno/Node.js backdoor. This backdoor enables credential theft, remote access, and delivery of additional payloads, posing significant risks to banks and public agencies.

1mo ago · hacker-news

A new modular malware named TELEPUZ has been spreading since April 2026 via ClickFix social engineering lures, which use clipboard hijacking (pastejacking) to trick users into executing malicious PowerShell commands. The malware, written in C, performs anti-analysis checks, disables security monitoring, and establishes persistent command-and-control (C2) communication using multiple fallback methods. It is capable of data theft, command execution, and browser manipulation via Chrome DevTools Protocol, and is likely offered as malware-as-a-service (MaaS) based on active development and distribution patterns.

1mo ago · hacker-news

Zoom has patched a critical vulnerability, CVE-2026-53412, in its Windows clients that could allow unauthenticated remote attackers to perform account takeover via improper input validation. Additionally, three high-severity vulnerabilities related to privilege escalation were addressed in various Zoom products for Windows. There is no evidence of active exploitation in the wild. Users are advised to update to the latest versions to mitigate these risks.

1mo ago · hacker-news

OpenAI has developed GPT-Red, an internal automated red-teaming model designed to identify and mitigate prompt injection vulnerabilities in its GPT series of models. GPT-Red simulates adversarial attacks by iterating through prompts to achieve malicious goals such as exfiltrating sensitive data or injecting malicious scripts, thereby improving the robustness of GPT-5.6 Sol. The model is used in adversarial training to harden production models against both direct and indirect prompt injections, significantly reducing failure rates compared to previous versions.

1mo ago · hacker-news

TuxBot v3 Evolution is a newly identified IoT botnet framework showing signs of large language model (LLM)-assisted development, though with functional flaws due to incomplete code. The malware targets IoT devices using brute-force attacks and known vulnerabilities, featuring a modular architecture with multiple C2 mechanisms including encrypted TCP, DGA, IRC, DNS, and P2P. It is attributed to the Keksec ecosystem based on shared infrastructure with Kaitori v3.9 and AISURU, indicating it is part of a broader portfolio of IoT botnets.

1mo ago · hacker-news

OkoBot is a malware framework targeting Windows users, active since April 2025, that injects phishing pages into legitimate cryptocurrency wallet applications like Ledger Live and Trezor Suite to steal recovery phrases. One of its modules, SeedHunter, hooks into Electron-based apps and waits for hardware wallet connections before displaying a malicious recovery page. The framework uses trojanized software and phishing lures to gain access, establishes persistent remote access via SSH and RDP, and deploys multiple surveillance and data-stealing plugins. Kaspersky attributes the campaign to an unknown actor but notes Russian-language artifacts and targeting patterns.

1mo ago · hacker-news

A critical vulnerability in the Cursor IDE on Windows allows malicious cloned repositories to trigger arbitrary code execution by placing a file named git.exe in the project root. When the repository is opened, Cursor automatically executes this binary without user consent, enabling attackers to run code with the user's privileges, including access to SSH keys and cloud tokens. Despite being reported in December 2025, no patch has been released, and the issue remains unaddressed in the latest version. Similar behavior has been observed in other AI-powered development tools, indicating a broader trend in untrusted search path vulnerabilities.

1mo ago · hacker-news

The article discusses the 'Approval Gap' in AI-era ad tech, where approved marketing tags can dynamically load unvetted fourth-party scripts, creating client-side security risks. These scripts operate with the same privileges as first-party code, potentially exposing sensitive customer data. The threat is exacerbated by AI-driven ad tech that rapidly evolves, making point-in-time security reviews insufficient. The piece emphasizes the need for continuous monitoring of digital supply chains to prevent unauthorized data access.

1mo ago · hacker-news

Security researcher Chaotic Eclipse has released a proof-of-concept (PoC) exploit named LegacyHive, which exploits a Windows User Profile Service vulnerability to load arbitrary hives and achieve privilege escalation. The exploit works on all supported Windows versions, including those updated with the July 2026 Patch Tuesday. The researcher claims the original version did not require additional credentials and could target any registry hive, raising concerns about potential misuse. This disclosure follows an ongoing dispute between the researcher and Microsoft over responsible vulnerability disclosure.

1mo ago · hacker-news

The article discusses how traditional SASE (Secure Access Service Edge) architectures are failing to keep pace with modern enterprise workflows involving SaaS applications, browsers, and generative AI tools. Due to encryption protocols like TLS 1.3 and HTTP/3, cloud proxies can no longer inspect traffic effectively, forcing organizations to bypass security for performance, creating blind spots. AI-driven workflows and autonomous agents now operate beyond network inspection, enabling potential data exfiltration through actions like pasting sensitive data into public LLMs. The solution proposed is shifting enforcement to the endpoint level, inspecting data at the point of interaction before it leaves the device.

1mo ago · hacker-news

Multiple critical vulnerabilities have been identified and patched in Firefox, Chrome, Adobe products, and VMware Avi Load Balancer. Mozilla warned of publicly available exploit code for two Firefox flaws, though no active exploitation has been observed. Adobe addressed 88 vulnerabilities across ColdFusion, Commerce, Experience Manager, and Illustrator, with several allowing arbitrary code execution. Broadcom patched a critical authentication bypass in VMware Avi Load Balancer that could allow network-accessible attackers to access the control plane.

1mo ago · hacker-news

SonicWall has identified active exploitation of two zero-day vulnerabilities in its SMA 1000 series appliances. CVE-2026-15409 is a critical SSRF flaw allowing unauthenticated remote attackers to manipulate internal requests, while CVE-2026-15410 is a post-authentication code injection vulnerability enabling arbitrary command execution with administrator privileges. The U.S. CISA has added both vulnerabilities to its Known Exploited Vulnerabilities catalog, mandating federal agencies to patch by July 17, 2026. Indicators of compromise include suspicious log entries and unexpected configuration changes, prompting forensic analysis and system remediation.

1mo ago · hacker-news

Four compromised npm packages under the @asyncapi namespace have been used to distribute a multi-stage botnet loader that downloads the Miasma malware from IPFS. The malicious code executes when the package is loaded via require(), not during install, evading traditional detection. The malware supports multiple C2 channels, enables credential theft, lateral movement, and includes a dead man's switch. The attack leveraged compromised CI/CD pipelines with legitimate OIDC attestations, not stolen npm tokens.

1mo ago · hacker-news

Microsoft's July 2026 Patch Tuesday addresses a record 622 vulnerabilities, including two actively exploited zero-days. CVE-2026-56164 is a privilege escalation flaw in SharePoint Server exploited by unauthenticated attackers over the network. CVE-2026-56155 is a local privilege escalation in Active Directory Federation Services exploited by authenticated attackers. Despite moderate severity ratings, both are under active attack and require immediate patching.

1mo ago · hacker-news

Researchers have identified a vulnerability in the Claude for Chrome browser extension that allows rogue extensions to trigger unauthorized actions, such as reading Gmail, Google Docs, and Calendar data, by exploiting forged script-generated clicks. The flaw stems from the extension's failure to validate the authenticity of user interactions (event.isTrusted), enabling malicious extensions to simulate clicks and initiate sensitive tasks. If users have enabled 'Act without asking' mode, these actions execute silently without approval, increasing the risk to critical levels. The issue remains unpatched as of version 1.0.80, despite being reported to Anthropic in May 2026.

1mo ago · hacker-news

SAP has released security updates to address multiple critical vulnerabilities, including CVE-2026-44747, a CVSS 9.9 out-of-bounds write flaw in SAP NetWeaver ABAP that could allow authenticated attackers to cause memory corruption and potentially access or modify sensitive data. Two other critical flaws were also patched: CVE-2026-27690, an HTTP smuggling vulnerability in SAP Approuter, and CVE-2026-44761, a default credentials issue in SAP Commerce Cloud stemming from sample configuration scripts. Although no active exploitation has been observed, attackers could leverage these flaws to gain unauthorized access, manipulate data, or cause denial-of-service conditions if left unpatched.

1mo ago · hacker-news

Two critical vulnerabilities in RabbitMQ, CVE-2026-57219 and CVE-2026-57221, could allow unauthenticated attackers to leak OAuth client secrets and enable authenticated users to bypass tenant boundaries by accessing cross-tenant queue metadata. The flaws, present since early 2024, affect RabbitMQ versions 3.13.0 and later and have been patched in recent releases. CVE-2026-57219 exposes a misconfigured HTTP API endpoint that leaks sensitive OAuth secrets, posing high risk in cloud or multi-tenant environments with exposed management interfaces.