CVE-2026-49163, CVE-2026-56161, CVE-2026-56181, CVE-2026-62836, CVE-2026-62869, CVE-2026-63913, CVE-2026-64561, CVE-2026-68823
Get tomorrow's brief in your inbox
Today: Malware can silently abuse Windows Hello keys for persistent Entra ID access without triggering prompts. NatJack attacks manipulate NAT tables to hijack TCP sessions across Windows and Linux. Microsoft patched critical Azure vulnerabilities while ransomware groups continue targeting healthcare and financial sectors.
Malware Can Abuse Windows Hello for Business Keys for Persistent Entra ID Access
Researcher Dirk-jan Mollema demonstrated that malware running in a signed-in Windows session can silently use the victim's Windows Hello for Business key to authenticate to Microsoft Entra ID without triggering biometric prompts or extracting private keys. The attacker can establish longer-term cloud access, register a controlled device, obtain a Primary Refresh Token valid for 90 days, and add authentication methods where policies permit. On TPM-backed systems, Windows ticketing keeps private-key operations available while the user is signed in, allowing code running as the user to request signed authentication data. The technique treats Windows Hello for Business keys as FIDO2 passkeys through WebAuthn, and the five-minute Entra ID challenge is not bound to a session, user, or tenant.
NatJack Attacks Hijack TCP Sessions and Spoof DNS by Manipulating NAT Tables
Security researcher Malcolm Stagg disclosed NatJack, an attack class that manipulates network address translation connection state to hijack active TCP sessions, spoof DNS responses, expose mapped ports, and exhaust NAT tables. Presented at Black Hat USA 2026, the research found affected behavior across Windows and Linux implementations. CVE-2026-56181 (CVSS 8.3, EPSS 0.002/15th percentile) affects Windows NAT used by Hyper-V. CVE-2026-63913 (CVSS 8.2, EPSS 0.006/46th percentile) affects Linux Netfilter conntrack. The attack generally requires privileged access to a system behind the same NAT as the victim. There is no single patch for the broader attack class.
26 claims tracked across 4 groups in the last 48 hours. These are unverified claims from ransomware leak sites, not confirmed breaches.
| Group | Victim | Sector | Country |
|---|---|---|---|
| Storm | Pioneer Bank | FinTech | United States |
| Storm | NCA Alarms | Security Systems | United States |
| Storm | Nelson Manufacturing | Aerospace Manufacturing | United States |
| Storm | Southern Indiana Radiological Associates | Healthcare | United States |
| Storm | Liberty Healthcare Corporation | Healthcare | United States |
| L Group | automobile-mueller.info | Automotive | Unknown |
| L Group | Universidade Veiga de Almeida | Education | Brazil |
| L Group | Rose Kennedy Greenway | Non-profit | United States |
| L Group | NoKota Packers | Food Production | United States |
| L Group | Coast Produce Company | Food Distribution | United States |
| Helix | Venture Logistics | Logistics | Unknown |
| Helix | Uber | Technology | United States |
| Helix | Highwoods Properties | Real Estate | United States |
| Helix | Morguard | Real Estate | Canada |
| Helix | Westland Insurance | Insurance | Canada |
| Barracuda | Namyang Industrial / Namyang Nexmo | Manufacturing | South Korea |
UNC6671 Rebrands: Multi-Brand Vishing Extortion Targets Financial Services and Enterprise Cloud
Google Threat Intelligence Group reports UNC6671 continues conducting data theft extortion despite the alleged retirement of the BlackFile extortion brand in May 2026. Telemetry and infrastructure analysis reveal the group has diversified operations across multiple extortion fronts including Redact, Pink, Helix, and Falcon. UNC6671 continues to rely on voice phishing and targets financial services and enterprise cloud environments.
Belarusian Ransom Cartel Mastermind Gets 16 Years in Prison
Maksim Silnikau, 40, the Belarusian creator and administrator of the Ransom Cartel ransomware, was sentenced to 16 years in prison in the US. Silnikau was extradited from Poland to the US in August 2024.
Cardiology Associates of Port Huron Allegedly Hacked in June
Cardiology Associates of Port Huron remains silent although they were allegedly hacked and had patient data stolen in June by the Orova ransomware group. The group is one of approximately 4 dozen new threat actor groups targeting US medical entities in the first half of 2026.
Dutch Retailers Bol and De Bijenkorf Warn of Data Breach
Online retailer Bol warned customers about a data breach involving one of its logistics partners. Unauthorized parties accessed the partner's systems, and some customer information may have been viewed or copied. Bol's own systems were not affected. Leaked data has appeared on dark web forums.
TeamPCP Linked To Redis Attacks Dating Back To 2020
Oligo Security research uncovered that TeamPCP has been active since 2020, compromising internet-facing infrastructure years before targeting the software supply chain. The connection is supported by overlapping domains, malware deployment paths, staging techniques, and operational tradecraft. The group has targeted Ray, Docker, Redis, and React environments. Recent campaigns include ShadowRay 2.0 (hijacking AI infrastructure into self-propagating botnets) and TA-NATALSTATUS (targeting exposed Redis servers to deliver cryptocurrency miners). New malware variants observed as recently as March 2026 include wiper functionality that checks for Iran timezone and executes destructive code.
ChainDrop: Self-Propagating npm Worm
Palo Alto Unit 42 published analysis of ChainDrop, an npm supply chain worm that extracts GitHub Actions runner secrets and uses Ethereum smart contracts for command-and-control routing.
Canadian Man Pleads Guilty in Snowflake Extortions
Connor Riley Moucka, 26, of Kitchener, Ontario, pleaded guilty to computer fraud and conspiracy to hack and extort more than 165 organizations that used Snowflake cloud data storage. Moucka also admitted to stealing call and text history records of more than 100 million AT&T customers.
CVE-2026-62869 Azure Entra ID Spoofing Vulnerability
Insufficient verification of data authenticity in Azure Entra ID allows an authorized attacker to perform spoofing over a network. No additional details or CVSS score available in Microsoft's advisory.
CVE-2026-68823 Azure Confidential Ledger Remote Code Execution
Exposed dangerous method or function in Azure Confidential Ledger allows an authorized attacker to execute code over a network. No additional details or CVSS score available.
CVE-2026-62836 Azure SQL Managed Instance Elevation of Privilege
Improper restriction of communication channel to intended endpoints in Azure SQL Managed Instance allows an unauthorized attacker to elevate privileges over a network.
CVE-2026-49163 Application Insights Profiler Elevation of Privilege
Improper limitation of a pathname to a restricted directory (path traversal) in Application Insights Profiler allows an authorized attacker to elevate privileges over a network.
CVE-2026-56161 Azure Logic Apps Information Disclosure
Improper access control in Azure Logic Apps allows an authorized attacker to disclose information over a network.
AI Recommendation Poisoning: How "Ask AI" Buttons Silently Alter LLM Memory
Microsoft Security catalogued AI Recommendation Poisoning in February 2026, identifying 31 companies across 14 industries deploying it, with more than 50 distinct prompts observed. The technique manipulates long-term memory in ChatGPT, Claude, Gemini, and Grok by embedding hidden prompt injection payloads inside "Ask AI" buttons on marketing and competitor comparison pages. When a logged-in user clicks the button, a pre-filled query executes immediately with no confirmation, instructing the AI to permanently save the vendor's domain as a "trusted source." The technique is formally tracked in MITRE ATLAS as AML.T0080 (Memory Poisoning), related to AML.T0051 (LLM Prompt Injection).
Claude Code and Gemini CLI Flaws Let GitHub Issue Reach CI Workflow Secrets
A GitHub issue opened by an account with no repository privileges was enough to execute code on the CI runners behind Anthropic's and Google's own coding-agent repositories. Novee Security ran the attack against each vendor's agent in the default configuration and presented the work at Black Hat USA on August 5. On OpenAI's repository, the attack was enough to hijack the next agent run.
Interrupt Injection Attack Can Bypass Spectre v2 Defenses on Intel and AMD CPUs
MIT CSAIL researchers Daniël Trujillo and Mengjia Yan disclosed INTERRUPT INJECTION, a technique where an unprivileged Linux program can time a hardware interrupt to land in the gap between a processor sanitizing its branch predictor and the kernel using it. On an AMD Zen 2 machine running Linux 6.14 with every default Spectre v2 mitigation enabled, the researchers demonstrated successful branch predictor re-poisoning after defenses ran.
Apple iCloud Private Relay Can Expose Real IPs Through WebKit Proxy Bypasses
Cybersecurity researchers disclosed a security issue with Apple's iCloud Private Relay tool that can expose a user's real IP address. iCloud Private Relay employs a dual-hop architecture to ensure users' privacy by routing Safari web traffic through two relays so that no single third-party, including Apple, can determine where the request originates.
CryptoJS Weak RNG Behind $5.7 Million in Crypto Wallet Drains
Coinspect identified CryptoJS.lib.WordArray.random() as the weak random number generator behind the Ill Bloom wallet drains. The function, introduced in the JavaScript cryptography library 12 years ago, supplied weak entropy that affected wallet apps used to generate recovery phrases. On-chain analysis puts measured theft across two sweeps since late May at a lower bound of $5.7 million across five crypto wallet apps.
Over 4,400 Rockwell PLCs Exposed Online, 22 Found in Water Attack Cities
Forescout found 22 internet-facing Rockwell Automation PLCs in cities hit by recent cyberattacks on US water utilities. Nineteen used the same mobile carrier network. An August 3 scan counted 4,407 exposed Rockwell controllers worldwide, including 2,844 in the United States. Forescout could not confirm any were compromised.
Truck Brake Controller's Safety Recall Doubled as Hidden Security Fix
NMFTA research shows a Bendix EC80 brake controller safety recall also patched remote code execution and denial-of-service vulnerabilities.
CVE-2026-64561 Zapscape KVM Flaw Could Let Privileged L1 Guest Code Escape to Linux Hosts
Zapscape, a new Linux kernel vulnerability, could allow an attacker with kernel privileges inside an L1 guest virtual machine to escape KVM isolation and execute code on the host. The risk applies when nested virtualization is exposed to untrusted guests. The flaw affects KVM/x86's shadow memory management unit.
Cisco Patches 12 SD-WAN and IOS XE Flaws, Including Three 9.9 CVSS Score Bugs
Cisco released updates to address multiple critical security vulnerabilities impacting Catalyst SD-WAN and IOS XE Software as part of a comprehensive internal security review. The security issues affect Cisco Catalyst SD-WAN Software regardless of device configuration, and Cisco IOS XE Software when running in autonomous or controller mode.
CISA ICS Advisories
Three new ICS advisories published: ICSMA-26-218-01 (Medixant RadiAnt DICOM out-of-bounds write, CVSS 4.3), ICSA-26-218-02 (Johnson Controls TL280 broken cryptographic algorithm, CVSS 4.1), and ICSA-26-218-01 (ABB Ability Zenon multiple vulnerabilities, CVSS 7.8).
Windows 10 LTSC 2021 Has Five Months Before Security Updates Cost Extra
Windows 10 LTSC 2021 Extended Security Update meter starts running in January 2027, unless you're on the IoT edition.
Multiple research presentations from Black Hat USA 2026 disclosed fundamental weaknesses in virtualization, NAT implementations, and AI coding agents. The Windows Hello for Business technique and NAT manipulation attacks both exploit legitimate features rather than implementation bugs, highlighting architectural security challenges. Ransomware groups continue rebranding and diversifying operations, with UNC6671 operating across multiple extortion brands despite announced retirement. Healthcare and financial services remain primary targets for both traditional ransomware and emerging vishing extortion campaigns.