← Carolina Clear Tech

Cyber Threat Brief

2026-11-09

Listen to this brief (13:49)

Download MP3
Show Notes

Show Notes - 2026-11-09

Stories Covered

CVEs Referenced

CVE-2026-107154, CVE-2026-107155, CVE-2026-109593, CVE-2026-109594, CVE-2026-123158, CVE-2026-19868, CVE-2026-82308

Indicators of Compromise

IP Addresses: 156.238.252.37

Read the full brief

Get tomorrow's brief in your inbox

Protect Your Business

Need a security assessment? See our cybersecurity packages.

View Services

Today: CISA added SonicOS MFA-bypass vulnerability CVE-2026-19868 to KEV, with a November 20 remediation deadline for federal civilian agencies. LNbits has instructed operators to shut down servers amid attacks; verify the project's current recovery guidance before restarting. Refrigeration-controller attacks are disrupting cold chains, while SANS has observed SSH exploitation attempts involving CVE-2026-123158.

Critical Alerts

SonicOS SSL-VPN Authentication Bypass Added to KEV (CVE-2026-19868)

Summary: CISA added CVE-2026-19868 to KEV after reports of possible exploitation; SonicWall says it investigated a report indicating potential exploitation. The high-severity flaw, listed with CVSS scores of 7.1 / 7.5, lets users with valid credentials for an alternate authentication method bypass SAML controls, potentially including SAML-enforced MFA. Enrichment lists EPSS at 0.006, or 0.6%, in the 50th percentile, and a November 20, 2026 federal remediation deadline; the low predictive score does not override KEV prioritization.

LNbits Calls for Server Shutdown Amid Wallet Attacks

Summary: LNbits instructed all operators to shut down their servers while it prepared security fixes, expanding an earlier warning to disable the Paywall extension. Operator Kevin Ravensberg said attackers withdrew 2,285,545 satoshis, approximately 0.023 bitcoin, from his server; the exploited extension was not publicly identified in the account. As of the article's 09:00 UTC update, no fixed release had been issued, and the developer's 24-to-48-hour patch estimate was not a confirmed release time.

SANS Observes SSH Exploitation Attempts (CVE-2026-123158)

Summary: SANS Internet Storm Center observed exploitation attempts against the "sshole" vulnerability, CVE-2026-123158, with the first identified attempt arriving November 5 using an all-"A" password. Its Cowrie honeypots record only the first 255 password characters, limiting analysis; earlier long-password submissions included misplaced shell commands rather than identified exploit payloads. Enrichment gives EPSS as 0.010, or 1.0%, in the 63rd percentile, but no CVSS score or affected implementation is established here.

IOCs & Detection

Unclassified SSH Download Payload

Summary: SANS documented a command submitted in an SSH password field that attempted to download an executable, save it as /tmp/b, and make it executable. The observation supports investigation of a download indicator, but does not establish a malware family, successful execution, or a connection to the identified sshole exploit attempts.

Business & Infrastructure Threats

Refrigeration Attacks Put Inventory and Equipment at Risk

Summary: CISA, FBI, HHS, USDA, and DOT warned that attackers are targeting industrial refrigeration and cold-chain monitoring systems, altering temperature and defrost settings or disrupting visibility. The article describes attacks affecting hundreds of European supermarkets, including small grocers and butchers; the French Blood Establishment confirmed extreme temperature and defrost changes at three plasma-storage facilities following an incident affecting an equipment provider. Attribution remains unresolved, and NoName057(16)'s responsibility claim is not established.

Frontline Education Breach Affected More Than 3 Million People

Summary: Frontline Education's Indiana filing puts its August breach at 3,085,266 affected individuals, substantially above earlier state filings totaling roughly 1.3 million. Frontline says attackers exploited a third-party software vulnerability and stole employee and student information from Frontline Central and School Health Management; potentially exposed fields vary by person and include Social Security numbers, government identifiers, and health insurance information. The company detected the intrusion August 14 and says it contained it the following day.

BOSSAI Operation Used AI Agents Against Exposed Services

Summary: Acronis researchers attribute an operation targeting hundreds of organizations in at least 11 countries to a single Taiwan-based attacker using AI agents. They describe access or data collection at roughly 10 organizations, including emergency-dispatch records and SQL-injection data theft; their detailed account distinguishes an exposed Taipei patient-data platform from confirmed theft of its records. The operation combined manually coordinated agents and locally operated models with safety training removed, rather than fully autonomous collaboration.

Windows / AD Security

BluePhishing Demonstrates Entra Agent-Blueprint Consent Abuse

Summary: Proofpoint's research-based BluePhishing proof of concept extends OAuth consent phishing to Microsoft Entra agent blueprints; Proofpoint has not identified in-the-wild activity. A victim's consent creates a blueprint principal that receives AgentIdentity.CreateAsManager, enabling tenant-local agent creation and operation with granted permissions. Proofpoint cites Microsoft announcement MC1506733 for default-user blueprint creation beginning in November 2026, making tenant configuration and consent governance immediate review priorities.

General Security News

NoName057(16) Publishes Personal Information of European Officials

Summary: Risky Biz reports that NoName057(16) published personal information concerning 18 high-profile European politicians and law-enforcement officials, with three German journalists also covered by the dox. The publication used a portal soliciting personal information about anti-Russia politicians and people involved in infrastructure takedowns. The outlet's suggestions about implied violence are commentary, not evidence of a specific planned attack.

RSA-1024 Research Reduces Certain Attack Costs

Summary: Google DeepMind says supervised AI agents improved attacks against obsolete RSA-1024, including a sandboxed signature forgery requiring 391 CPU core-years and roughly four billion advance signing queries. The demonstrated attack did not recover the private key and does not work against devices that hash messages internally; a separate multi-key factoring improvement remains theoretical and unimplemented. DeepMind says modern protocols using RSA-2048 or larger remain secure against these results.

Patch Priority

Vulnerability Disclosures

Additional SonicOS Fixes and End-of-Life GMS Exposure

Summary: SonicWall also patched management-interface session-validation flaw CVE-2026-107154 and permissive cross-domain-policy flaw CVE-2026-107155, both CVSS 8.8, plus SSL-VPN heap-buffer-overflow flaw CVE-2026-82308, CVSS 7.5. Their enriched EPSS values are 0.007, 0.008, and 0.006 respectively. Separately, SonicWall described CVE-2026-109593 and CVE-2026-109594 as unpatched vulnerabilities in end-of-life Global Management System deployments and urged migration to Network Security Manager.

Trends & Context

Internet-accessible administrative services, alternate authentication paths, and consent-based identities are recurring access risks across these stories. Prioritize observed exploitation and KEV status over low EPSS values, while keeping research demonstrations, exposed assets, and confirmed intrusions distinct.