‼️ GitLab has patched a critical 9.9-rated flaw in its self-hosted AI Gateway that could let low-privileged Duo Agent users escape the prompt sandbox and execute commands on the underlying server.

Affected self-hosted deployments should update immediately.

Read: https://thehackernews.com/2026/10/gitlab-patches-critical-self-hosted-ai.html

⚠️ Antino turns Outlook and OneDrive into command-and-control channels.

China-nexus UAT-11587 is using the Rust backdoor in spear-phishing attacks targeting government and policy organizations across Asia.

How the attack chain works: https://thehackernews.com/2026/10/antino-backdoor-uses-outlook-and.html

Visibility alone is no longer enough.

A new report brings together 10 security companies on what comes next: continuous identity governance, telemetry that can be reused, enforced remediation, and AI that accelerates investigations without replacing human judgment.

Full report: https://thehackernews.com/2026/10/the-state-of-cybersecurity-in-2026key.html

⚠️ Unpatched SharePoint is still getting orgs ransomed.

Warlock hit a water utility, a telecom, a government body, and a university in the last 2 months. They killed security tools on 40 hosts in ~2 hours, then used SYSVOL to push ransomware to 33 machines.

Read: https://thehackernews.com/2026/10/warlock-exploits-sharepoint-flaws-to.html

⚡ MI5 named a Chinese institute an MSS front and said 100+ UK-linked academics helped fund its spy tech.

A 30 Sept alert says CGTRI exists to bankroll work that improves MSS capability: AI, cyber, covert comms, steganography. Some academics may not know who paid.

Beijing calls it fabricated.

Read: https://thehackernews.com/2026/10/mi5-says-chinas-mss-funded-research.html

TA419 is targeting U.S. AI policy experts with Microsoft AitM phishing.

The China-aligned group waits for targets to reply, then sends a shortened link to a Frameless BitB page built to capture credentials and session cookies.

Read about the technique: https://thehackernews.com/2026/10/china-aligned-ta419-targets-us-ai.html

‼️ A suspected #ShinyHunters member is reportedly helping the FBI identify others in the group.

Rey, also known as ReyXBF, was allegedly detained in Jordan on September 29. His cooperation is now feeding an ongoing effort to pursue more members.

Read: https://thehackernews.com/2026/10/shinyhunters-suspect-rey-reportedly.html

‼️ Citrix has patched a new NetScaler zero-day exploited in targeted attacks.

CVE-2026-88779 affects certain SAML-configured deployments & can cause denial-of-service, with repeated triggering potentially leaving services unavailable.

Details - https://thehackernews.com/2026/10/new-netscaler-zero-day-exploited-in.html

An AI pentesting agent says it exploited a bug. Did it?

XRanges for AI watches from inside the target, recording exploit-chain progress, coverage, boundary violations, and whether the environment survives the run.

Inside the scoring: https://thehackernews.com/2026/09/545-hackers-tested-it-first-now-xranges.html

⚠️ Attackers are targeting a Rejetto HFS flaw that enables forged admin sessions and remote code execution.

CVE-2026-61500 affects HFS 3.0.0–3.2.0. VulnCheck detected exploitation attempts against vulnerable U.S. hosts after a public PoC was released.

Read - https://thehackernews.com/2026/10/attackers-target-rejetto-hfs-flaw-that.html

A financial app treated a GUID like proof of access.

In a penetration test described by Sygnia's Zach Mead, the AI-assisted onboarding flow could issue or restore an applicant token if another applicant’s GUID was obtained, potentially exposing sensitive personal and financial data.

How the trust failure worked: https://thehackernews.com/expert-insights/2026/10/when-ai-writes-code-who-owns-security.html

⚡ Apple is tightening macOS Full Disk Access, a permission that can let AI agents read files, mail, messages, browsing history, and more.

Future access will require explicit user action.

Here's what Apple is changing: https://thehackernews.com/2026/10/apple-plans-tighter-macos-full-disk.html

🚨 Cling hides C2 commands in STUN traffic as threat actors were observed attempting to exploit Realtek Jungle SDK flaw CVE-2021-35394. A subset of that activity delivered the botnet.

Read how Cling uses STUN for command-and-control, persistence, propagation, proxying, tunneling, and DoS: https://thehackernews.com/2026/10/realtek-jungle-sdk-exploit-attempts.html

CJIS compliance starts with stronger authentication.

Weak passwords and inconsistent MFA enforcement can create compliance and security challenges for agencies and organizations that handle criminal justice information.

Download our practical guide to learn:

✅ Key CJIS password requirements
✅ MFA compliance best practices
✅ Common compliance gaps to avoid
✅ Steps to strengthen your security posture

Get the guide: https://thn.news/password-mfa-standards

15 cyber stories worth 2 minutes of your attention:

• NetScaler + FortiMail 0-days
• Spectre v2 leaks root hashes
• CosmicPulse phishing
• NeedyMantis persistence
• RatHat + Gemini targeting
• 13K AI-leaked screenshots
• Ransomware arrests
• Microsoft X hijack
• WordPress credential theft
• PageBreak AI vuln hunting
• Milk Dragon phishing
• NodeStealer upgrades
• Direct Send bypass
• PaperCut exploitation
• AI models building exploits
• Plus a huge CVE pile

Full ThreatsDay recap: https://thehackernews.com/2026/10/weekly-recap-netscaler-and-fortimail-0.html

GitGuardian found 28.65 million new hardcoded secrets in public GitHub commits in 2025, up 34% year over year.

It also found 24,008 unique secrets in public MCP configuration files, including 2,117 verified as valid.

See how credentials are spreading across code, endpoints, and AI tooling: https://thehackernews.com/2026/10/the-credential-layer-is-expanding.html

‼️ ALERT - Exchange admins should review this now.

CVE-2026-96940 can allow an authenticated attacker to access other users’ mailboxes and read emails and attachments within the same organization.

Microsoft has issued out-of-band fixes.

Read: https://thehackernews.com/2026/10/microsoft-exchange-flaw-lets.html

🚨 ClickFix has a new trick... the malicious payload is already sitting in your browser cache before you paste the command.

Compromised sites preload scripts disguised as PNGs, letting pasted commands bypass Windows Run's ~260-character limit and launch a multi-stage malware chain.

Read: https://thehackernews.com/2026/10/clickfix-smuggles-payloads-through.html

⚠️ Denmark says unauthorized parties accessed names, addresses and CPR numbers for about 8.8 million people, roughly 4 in 5 records in its national register, via a private company’s lawful access.

Automated lookups ran 10 days; police are investigating.

Details → https://thehackernews.com/2026/10/denmark-says-attackers-accessed-cpr.html

‼️ A security patch wasn’t applied. Thousands of FBI employees had personal details stolen.

The FBI has now removed an Accenture contractor over the security failure tied to the ShinyHunters breach.

Read the full report → https://thehackernews.com/2026/10/fbi-removes-accenture-contractor-after.html