← November 2026
News 2026-11-14

IAM & Security Weekly Briefing

Microsoft’s November Patch Tuesday addressed 1,259 flaws, including an exploited Outlook bug that may expose NTLMv2 hashes.

IAM & Security Weekly Briefing
Open report

IAM & Security Weekly Briefing

Week of: 2026-11-08 to 2026-11-14 Reporting window: Most recently completed Sunday–Saturday (excludes the in-progress week).


1. Executive Summary (TL;DR)

  • Microsoft’s November Patch Tuesday addressed 1,259 flaws, including an exploited Outlook bug that may expose NTLMv2 hashes.
  • Stolen credentials for 4,837 GitHub accounts were used to plant a CI secret stealer in at least 1,806 repositories.
  • A Google session hijacking campaign used calendar invites and live proxies to steal live corporate sessions, defeating normal link checks.
  • Cursor confirmed a breach after stolen Fireworks AI credentials enabled a malicious Claude Code extension in its marketplace.
  • N-able N-central flaws were exploited against 21 servers, while Check Point gateways were hit via PHP2Shell to deploy a backdoor and wipe ESXi files.
  • SilentAir intercepted SMS 2FA at Asian telecoms, and DPRK-linked FaceStealer used deepfake video calls to steal over $400 million.
  • Operation Endgame seized roughly 6.4 million stolen logins, underscoring how credential theft still fuels ransomware and account takeover.
  • AI agents and API keys were a recurring non-human identity risk, from Railway’s stolen AI keys to secret-leaking coding agents and exposed Grok-built apps.

2. Top IAM & Security News

Microsoft Patches 1,259 Flaws, Including Exploited Outlook Bug That May Expose NTLMv2 Hashes

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-11
  • What happened: Microsoft released patches for 1,259 vulnerabilities, including an exploited Outlook bug that may expose NTLMv2 hashes.
  • Why it matters: Exposed NTLMv2 hashes are credential material that can be used in relay or cracking attacks, so Outlook patching and NTLM reduction should be prioritized.

Attacker Used 4,837 Hijacked GitHub Accounts to Plant CI Secret Stealer in 1,806+ Repositories

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-12
  • What happened: An attacker used stolen credentials for 4,837 GitHub accounts to plant a secret stealer in repositories those accounts could write to, targeting secrets and cloud credentials reachable by CI workflows.
  • Why it matters: Compromised developer identities and CI secrets can pivot directly into cloud environments; audit GitHub account security, token scope, and repository write access.

Google Session Hijacking Campaign Used Calendar Invites and Live Proxies to Steal Corporate Sessions

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-12
  • What happened: Since mid-August 2026, a phishing operation has used calendar invites and live proxies to steal live Google sessions from corporate users, prompting takedown of hundreds of attacker domains under Operation CloudFire.
  • Why it matters: Session token theft bypasses password and MFA prompts; prioritize token protection, conditional access, and rapid session revocation.

Cursor Confirms Breach After Stolen Fireworks AI Credentials Led to Malicious Claude Code Extension

  • Source: BleepingComputer
  • Link: BleepingComputer
  • Date: 2026-11-14
  • What happened: Cursor confirmed a security breach that allowed hackers to distribute a malicious Claude Code extension through its marketplace after stealing credentials from AI infrastructure provider Fireworks AI.
  • Why it matters: Marketplace and AI tooling supply chains are now credential conduits; vet extensions, scope AI provider credentials, and monitor developer marketplace installs.

N-central Flaws Exploited to Infect 21 Servers, Including 19 Running Hotfix 2

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-13
  • What happened: Attackers exploited N-able N-central flaws to compromise 21 servers, including 19 running Hotfix 2, prompting N-able to restrict certain features and integrations.
  • Why it matters: RMM platforms are high-value access paths to managed environments; validate hotfix efficacy, restrict internet exposure, and monitor for rogue remote-management activity.

Check Point Gateways Exploited via PHP2Shell to Deploy TSGA Backdoor and Wipe ESXi Files

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-13
  • What happened: An unknown threat actor targeted unpatched Check Point Security Gateways, using PHP2Shell vulnerabilities and a Gaia OS privilege-escalation flaw to deploy the TSGA backdoor and wipe ESXi files, beginning October 30, 2026.
  • Why it matters: Compromised security gateways can provide trusted access and destructive reach; emergency patching and integrity checks of edge devices are essential.

SilentAir Hijacked Asian Telecom Networks to Intercept SMS 2FA Codes

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-13
  • What happened: Enea identified a state-level spying group, SilentAir, using hacked equipment inside Asian mobile operators to intercept calls and texts, including SMS 2FA codes, mostly targeting users in Taiwan.
  • Why it matters: SMS-based MFA remains interceptable at the carrier level; high-risk users should move to phishing-resistant authenticators.

FBI: North Korean FaceStealer Campaign Stole Over $400 Million in Crypto via Deepfake Video Calls

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-13
  • What happened: The FBI attributed more than $400 million in cryptocurrency theft over 18 months to DPRK-linked actors using a social engineering scheme codenamed FaceStealer, including deepfake video calls.
  • Why it matters: Deepfake-enhanced social engineering defeats voice and video identity verification; strengthen out-of-band verification and high-value transaction controls.

3. AI, Identity & Emerging Tech

Railway Says Stolen Customer AI Keys Led to $740,000 in Estimated Unauthorized Usage Charges

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-11
  • What happened: Railway disclosed that AI provider keys customers stored on its platform were stolen in an October breach and used for an estimated $740,000 in unauthorized spending, which the company said it would cover where providers do not refund.
  • Why it matters: AI API keys are high-value non-human identities; inventory, scope, rotate, and monitor them for anomalous spend.

AI Coding Agents Leak Thousands of Secrets to GitHub

  • Source: Dark Reading
  • Link: Dark Reading
  • Date: 2026-11-10
  • What happened: A command in Claude Code inadvertently saves files containing credentials and sensitive data to public Git repositories, creating exposure for organizations.
  • Why it matters: Agentic developer tools can create unintended data flows; enforce secret scanning, least-privilege repository access, and agent guardrails.

Microsoft Researchers Show AI Agents Can Be Backdoored by Editing Fewer Than 10 Neurons

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-11
  • What happened: Microsoft researchers demonstrated that editing fewer than 10 rarely used neurons can implant a hidden backdoor in an AI model, causing an agent to run an attacker’s command with its own permissions when triggered.
  • Why it matters: Model integrity is now an identity risk because compromised agents act with their assigned privileges; protect model weights and verify agent behavior.

Token Researchers Describe an AI Agent That Broke Their Own Security Rules

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-11
  • What happened: Token built a browser agent that broke the company’s three principles for securing AI agents—separate identity, job-appropriate access, and independent traceability—by default.
  • Why it matters: AI agents need dedicated identities, least-privilege access, and independent audit trails rather than inheriting a user’s authenticated browser session.

Over 35,000 Grok-Built Apps Found Exposed Online, Many Without Sign-In

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-11
  • What happened: Claroc found 37,571 Grok-built apps online, with 35,578 open to anyone without a sign-in and about one in ten containing internal business or confidential data.
  • Why it matters: AI-generated apps often ship without authentication; require identity controls and review before exposing them.

4. Cyber Threats & Attack Trends

Operation Endgame Takedown Seizes 6.4 Million Stolen Logins

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-13
  • What happened: An international police operation took down servers behind Gootloader, CastleRAT, and StealC and seized about 6.4 million stolen logins belonging to more than 40,000 victims, according to Germany’s BKA.
  • Why it matters: Credential theft remains the feedstock for ransomware; assume exposed credentials are in circulation and monitor for reuse.

Luna Moth Retargets Law Firms With Social Engineering Aimed at Microsoft Access Tokens

  • Source: Dark Reading
  • Link: Dark Reading
  • Date: 2026-11-12
  • What happened: The extortion group is retargeting law firms that refused to pay a first ransom, using a social engineering con aimed at stealing Microsoft access tokens.
  • Why it matters: Access-token theft can bypass password-based controls; protect token issuance, conditional access, and session revocation for high-risk users.

FBI Flash Alert: Artex Tool Automates Attacks From Targeting to Privileged Access in Under Three Hours

  • Source: The Record
  • Link: The Record
  • Date: 2026-11-13
  • What happened: The FBI warned that the open-source Artex hacking tool significantly reduces the time needed to chain attack techniques, with some intrusions reaching privileged access in under three hours.
  • Why it matters: AI-assisted attack automation compresses defender response windows; identity teams need faster detection and containment of credential abuse.

AntiFox Agentic Pentesting Platform Linked to AWS Credential Theft

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-11
  • What happened: Researchers disclosed that an emerging agentic penetration-testing platform called AntiFox was used to steal AWS credentials and pivot laterally in compromised environments.
  • Why it matters: Cloud credentials are the objective; enforce short-lived credentials, least-privilege roles, and monitoring for anomalous AWS API activity.

Trezor Confirms Hardware Implants in Three Wallets That Can Steal Recovery Phrases

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-13
  • What happened: Trezor confirmed that three Safe 5 wallets sold by reseller CryptoBilis contained hardware implants designed to copy recovery phrases and send them over cellular networks.
  • Why it matters: Recovery phrases are ultimate credentials; verify hardware supply chains and assume reseller-sourced devices may be tampered with.

186 Fake AI Domains Redirect Users to ClickFix Lures and Browser Hijackers

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-12
  • What happened: Researchers found 186 lookalike domains impersonating AI tools and serving ClickFix-style lures, browser-hijacking malware, and search-hijacking extensions.
  • Why it matters: Credential and session theft increasingly hides behind trusted AI brands; block lookalikes, train users, and monitor browser extensions.

5. Product Updates & Vendor News

Citrix Patches 12 NetScaler and Secure Access Client Flaws, Including Authentication Bypass

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-12
  • What happened: Citrix released patches for multiple critical flaws in NetScaler ADC, NetScaler Gateway, and NetScaler Console that could lead to denial-of-service or authentication bypass, including CVE-2026-120868 with a CVSS score of 9.3.
  • Why it matters: NetScaler is often the authentication and access gateway for critical applications; an auth bypass can expose internal sessions and credentials.

Critical Red Hat Quay Flaw Allowed Reading and Overwriting All Private Container Images

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-12
  • What happened: A critical Quay flaw let an attacker with registry access read and overwrite private container images in every repository; Quay enables anonymous access by default, and fixed builds were released for eight versions.
  • Why it matters: Container registries are a software-supply-chain trust boundary; anonymous access and weak repository permissions can let attackers poison images.

eMudhra Revokes 8 More Intermediate CAs Tied to October Key Compromise

  • Source: The Hacker News
  • Link: The Hacker News
  • Date: 2026-11-11
  • What happened: eMudhra revoked eight more intermediate certificate authorities after an in-house Python script issue tied to its October key compromise, even though the keys themselves were not weakened.
  • Why it matters: Certificate authorities underpin TLS and certificate-based identity; revocation events require trust-store and dependency review.

Apple Details Secure Exclaves for Isolating Camera and Microphone Data

  • Source: SecurityWeek
  • Link: SecurityWeek
  • Date: 2026-11-14
  • What happened: Apple detailed an architecture designed to isolate ambient sensor data from the main OS on HomeView and new iPhones.
  • Why it matters: Hardware-backed isolation can improve endpoint trust, but security teams should assess how platform changes affect sensor-data protection and compliance.

6. Practical Security Takeaways

  • Patch identity-facing and edge systems first—Outlook, Check Point gateways, N-central, and Citrix NetScaler—and verify hotfix efficacy rather than assuming patch level.
  • Audit GitHub accounts, personal access tokens, CI secrets, and repository write access; rotate exposed secrets and move to short-lived credentials.
  • Treat session tokens as credentials: enable token protection and conditional access, monitor for calendar/OAuth lures, and revoke sessions on suspicion.
  • Accelerate phishing-resistant MFA such as passkeys/FIDO2 for high-risk users and reduce reliance on SMS and voice-based verification.
  • Inventory and scope non-human identities, including AI provider keys, API keys, service accounts, and CI tokens; rotate and monitor usage.
  • Give AI agents dedicated identities, least-privilege access, human approval for sensitive actions, independent audit trails, and kill switches.
  • Remove anonymous or default access from container registries and management planes, and review repository and image permissions.
  • Reduce NTLM reliance and monitor for NTLMv2 hash exposure; prioritize credential hygiene after patch disclosures.
  • Review PKI trust stores and certificate dependencies after CA revocations and certificate-authority incidents.
  • Assume credentials seized in law-enforcement takedowns remain viable; force resets and monitor for credential reuse.

7. Trends to Watch

  • AI agents will become a formal non-human identity class, requiring dedicated credentials, least privilege, and independent audit trails.
  • Session-token theft and adversary-in-the-middle proxying will push enterprises toward token binding and phishing-resistant authentication.
  • AI-assisted attack tooling such as Artex and AntiFox will continue to compress the time from initial access to privileged access.
  • SMS 2FA and voice/video identity checks will keep losing trust as telecom interception and deepfake social engineering mature.
  • Developer tooling, CI/CD pipelines, RMM platforms, and edge gateways will remain priority access paths for attackers.

Sources

More from News