Daily Cybersecurity Briefing — September 23, 2026

Your daily roundup of the most critical cybersecurity developments from the past 24 hours.

ACTIVELY EXPLOITED VULNERABILITIES

• D-Link CVE-2026-86296 (CVSS: Maximum) — D-Link warned of a maximum-severity vulnerability in legacy DIR-822A dual-band Wi-Fi routers. Public PoC exploit code exists and no patch is available. Customers urged to retire and replace affected devices immediately.

• Linux Kernel (3 CVEs) — CISA warns of active exploitation of three Linux kernel vulnerabilities, one rated critical. FCEB agencies must remediate promptly per KEV guidance.

• Microsoft Patch Tuesday (September 2026) — 974 CVEs addressed including two actively exploited zero-days: CVE-2026-81963 and CVE-2026-85880. FCEB agencies must apply patches per KEV deadlines.

• WordPress 'Click2Shell' CSRF — Technical details and PoC published for a new CSRF vulnerability in WordPress Core. High exploitation potential; administrators should deploy WAF rules immediately.

• CISA KEV Additions (Sept 16) — CVE-2026-76460 (Cisco ISE – Incorrect Use of Privileged APIs) and CVE-2026-87886 (Acronis Backup – Incorrect Default Permissions) added to the KEV catalog. FCEB patch deadline applies.

BR3ACHES

• LastPass GitHub Spoof + BYOVD Infostealer (Sept 23) — Attackers spoofed the LastPass repository on GitHub, leveraged a Microsoft-signed driver to disable 145 security products (BYOVD technique), then deployed an infostealer. Full victim scope unconfirmed. Verify LastPass-related GitHub repositories and monitor endpoints for disabled security product alerts.

THREATS & POLICY

• Google Fined €403M by Irish DPC — Ireland's Data Protection Commission fined Google €403 million ($463M USD) for multiple GDPR violations related to improper processing of users' location data.

• White House EO on AI Cybersecurity (June 2026) — EO establishes voluntary framework for advanced AI developers to provide the federal government pre-release model access for national security and cybersecurity assessments.

• White House EO on Transnational Cybercrime (Aug 2026) — Directs agencies to treat cybercrime organizations as transnational criminal networks; AG directed to prioritize AI-enabled cybercrime enforcement.

• Iranian APT Targeting ICS Vendors (Ongoing) — US government confirmed Iranian threat actors targeting Siemens, Schneider Electric, and Rockwell Automation ICS devices in disruptive attacks on energy, water, and government critical infrastructure.

CLOUD & SAAS SECURITY

• Passkey-Themed Social Engineering → Cloud Compromise — Microsoft Security Blog (Sept 9) documented a campaign using passkey-themed phishing to bypass MFA and compromise identity and cloud resources. Attackers exploit user confusion around passkey adoption.

• IAM Misconfiguration Epidemic — Research finds 80-98% of cloud accounts (AWS, Azure, GCP) have weak IAM controls; over 50% of Azure Entra ID accounts lack MFA. Misconfigurations and unreviewed OAuth grants are the #1 source of SaaS data exposure. Identity compromise underpins 83% of cloud breaches.

IDENTITY & AUTHENTICATION

• MFA Bypass Techniques Active in 2026 — Adversaries exploit SIM swapping, MFA push bombing, and session cookie hijacking. Legacy protocols (IMAP/POP3) remain bypass vectors. OAuth token theft converts phished short-lived codes into long-lived tokens providing months of persistent access.

• Identity as the #1 Cloud Attack Vector — An attacker authenticated as a legitimate user bypasses nearly every downstream security control. Identity visibility is now foundational to any 2026 security program.

MOBILE SECURITY

• CISA Spyware Alert — CISA urged mobile security hardening against sophisticated spyware targeting journalists, activists, lawyers, and executives. Mercenary spyware vendors now deploy zero-click exploits requiring no user interaction.

• FBI Warning on China-Made Mobile Apps — FBI issued a formal warning about data security risks from mobile apps developed by Chinese-owned or affiliated companies.

• NGate Android Malware (New Variant) — A new NGate Android malware variant abuses the legitimate HandyPay app; researchers note the malicious code appears AI-generated, signaling AI-assisted mobile malware development by threat actors.

ICS/OT/IoT

• Water Treatment Facility Ransomware (Midwest US) — A Midwest water treatment facility suffered a ransomware attack temporarily disrupting operations. CISA issued an emergency directive for critical infrastructure operators with affected ICS systems.

• OT-Targeting Ransomware Variant — A new ransomware family specifically targeting OT environments in manufacturing was identified, exploiting known vulnerabilities in legacy ICS/SCADA systems.

• Siemens ICS Security Advisory — Siemens published an advisory detailing multiple vulnerabilities in ICS products enabling remote arbitrary code execution. Organizations should review and apply mitigations immediately.

• Iranian APT: Siemens, Schneider Electric, Rockwell Automation — US government confirmed Iranian threat actors targeting ICS devices from these three major vendors in disruptive attacks across energy, water, and government critical infrastructure.

RESEARCH & TOOLS

• Microsoft Patch Tuesday (September 2026) — Largest single patch release of 2026: 974 CVEs including two actively exploited zero-days (CVE-2026-81963 and CVE-2026-85880). Enterprise teams should prioritize Windows patching this week.

• Click2Shell WordPress CSRF PoC — Full technical details and working PoC exploit released for WordPress Core 'Click2Shell' CSRF vulnerability. Exploitation is trivial; patch immediately or deploy WAF mitigations.

• Recorded Future H1 2026 Malware & Vulnerability Trends — 1,328 actively exploited non-Microsoft CVEs tracked across 282 vendors; 243 tied to ransomware campaigns. AI-assisted frontier models are increasingly applied in security research.

AI SECURITY

• Prompt Injection Still #1 Agentic AI Threat — OWASP confirms prompt injection remains the top exploited vulnerability in production agentic AI systems. An injection that once embarrassed a chatbot can now execute code, exfiltrate data, or send emails autonomously. Current defenses are bypassed by adaptive attacks.

• EchoLeak – Zero-Click Prompt Injection in M365 Copilot — Researchers demonstrated a zero-click prompt injection in Microsoft 365 Copilot enabling silent access and exfiltration of enterprise data with no user interaction. All M365 Copilot deployments affected.

• CVE-2025-53773 – GitHub Copilot RCE via Prompt Injection (CVSS 9.6) — Hidden prompt injection in GitHub PR descriptions enables remote code execution via GitHub Copilot. Microsoft has patched this; all GitHub Copilot users must update immediately.

• Prompts as Shells – RCE in AI Agent Frameworks (May 2026) — Microsoft Security Blog documented vulnerabilities in AI agent frameworks where prompt inputs escalate to shell execution enabling infrastructure-level RCE. Over-privileged agentic AI deployments are highest risk.

• AI-Generated Malware in the Wild — The NGate Android malware variant and the Sept 23 LastPass GitHub spoof infostealer campaign both appear to leverage AI-generated code, confirming threat actors are incorporating AI tooling into malware development pipelines.

Sources

CISA Known Exploited Vulnerabilities Catalog: https://www.cisa.gov/known-exploited-vulnerabilities-catalog

CISA KEV Additions Sept 16: https://www.cisa.gov/news-events/alerts/2026/09/16/cisa-adds-two-known-exploited-vulnerabilities-catalog

CISA KEV Additions Sept 9: https://www.cisa.gov/news-events/alerts/2026/09/09/cisa-adds-four-known-exploited-vulnerabilities-catalog

Senserva KEV Tracker Sept 2026: https://senserva.com/exploited-this-week.html

SecurityWeek – Iranian Hackers Targeting Siemens, Schneider, Rockwell ICS: https://www.securityweek.com/us-warns-of-iranian-hackers-targeting-siemens-schneider-and-rockwell-ics-devices/

Microsoft Security Blog – Passkey-Themed Social Engineering: https://www.microsoft.com/en-us/security/blog/2026/09/09/passkey-themed-social-engineering-leads-identity-cloud-compromise/

Microsoft Security Blog – Prompts Become Shells (RCE in AI Agent Frameworks): https://www.microsoft.com/en-us/security/blog/2026/05/07/prompts-become-shells-rce-vulnerabilities-ai-agent-frameworks/

The Hacker News – Cloud Security Checklist: https://thehackernews.com/2026/09/your-cloud-security-checklist-doesnt.html

The Hacker News – Identity Visibility in 2026: https://thehackernews.com/2026/09/identity-visibility-in-2026-foundation.html

Help Net Security – OWASP Prompt Injection AI Failures: https://www.helpnetsecurity.com/2026/06/11/owasp-prompt-injection-ai-security-failures/

Forescout – ICS Cybersecurity in 2026: https://www.forescout.com/blog/ics-cybersecurity-in-2026-vulnerabilities-and-the-path-forward/

White House – Transnational Cybercrime EO (Aug 2026): https://www.whitehouse.gov/presidential-actions/2026/08/expanding-capabilities-to-combat-transnational-cyber-enabled-crime/

Recorded Future H1 2026 Malware Trends: https://www.recordedfuture.com/research/h1-2026-malware-vulnerability-trends

SecurityWeek – FBI Warns of China-Made Mobile App Risks: https://www.securityweek.com/fbi-warns-of-data-security-risks-from-china-made-mobile-apps/

Cybersecurity Dive – CISA Spyware Alert: https://www.cybersecuritydive.com/news/cisa-spyware-alert-messaging-apps-security-warning/806429/