Window: approximately 19 November 09:00 SGT to 20 November 09:00 SGT (24 hours). New and overnight developments, deduplicated against the 19 November edition, with gaps stated plainly.
After its 48-hour deadline expired, n0n published customer marketing data for all three on 19 November. JPMorgan, Visa and Mastercard each confirmed a third-party vendor incident; no payment data is in the published sets, and the Mastercard set looks like test data.
Google Threat Intelligence Group reports suspected PRC-nexus actor ARMORY CASTLE exploited CVE-2026-123890, an unauthenticated authentication bypass in vCenter Server, as a zero-day in early September. Broadcom patched it on 19 November (VMSA-2026-0009); there is no workaround.
PHP released 8.2.36, 8.3.37, 8.4.28, 8.5.13 and 8.6.0 on 19 November, fixing 17 CVEs, seven of them rated critical (9.2 CVSS 4.0), including session-handling remote code execution. No exploitation has been reported.
CISA added nothing to the Known Exploited Vulnerabilities catalogue on 19 November; it remains at version 2026.11.18 with 1,786 entries (verified against the live feed in this run). Synology, Apache Axis and REDCap items added 18 November fall due 21 November.
n0n published Visa, JPMorgan Chase and Mastercard customer data after its deadline — and a VMware vCenter zero-day exploited since September was patched. The n0n group published marketing records for all three firms on 19 November after the 48-hour deadline set on 16 November expired. JPMorgan confirmed third-party service providers were targeted and marketing data that may include customer names and email addresses was accessed; Visa confirmed an incident at an Asia-Pacific third-party provider; Mastercard confirmed a limited exposure from one benefit programme. No card numbers, passwords or balances are in the published sets, and the Mastercard set — about 2,000 poorly formatted entries — looks like test data rather than genuine customer records. Separately, Google Threat Intelligence Group disclosed that suspected PRC-nexus actor ARMORY CASTLE exploited CVE-2026-123890, a critical unauthenticated authentication bypass in VMware vCenter Server (9.8), as a zero-day in early September, stealing the SAML signing key; Broadcom patched it on 19 November with no workaround. CISA added nothing new to KEV on 19 November.
Third-party risk keeps dominating: a bank’s own digital-banking middleware was turned into a backdoor, and payout APIs were looted in India. Forescout disclosed “BankWeaver” on 19 November: attackers abused a misconfigured Temenos Quantum Fabric deployment at a financial institution, exchanging an anonymous token for an admin token without credentials and planting web shells; six more banks share the same exposure, though their compromise is not confirmed. In India, Gurugram police arrested two men over an attack on PayU Payments’ cloud servers that manipulated its payout API to siphon ₹25.67 crore through 265 unauthorised transactions. KPMG Canada confirmed unauthorised access to some clients’ tax documents, Vietnam’s National Credit Information Center was hacked, Bell Financial said some clients’ data was likely accessed, and Plaid’s KARTE disclosures now total more than 84 million records after DHC and Sharp published their counts. Regulators responded: Japan’s FSA issued a reporting order to Rakuten Securities, and Ireland’s minister confirmed the Wellington IT vulnerability is closed with no evidence of members’ funds being accessed.
Patch PHP and Splunk as well as vCenter — and the European Commission confirmed an AI-assisted attack. PHP’s 19 November release fixes seven critical remote-code-execution flaws (9.2 CVSS 4.0), and Splunk’s November update fixes 146 flaws including an unauthenticated authentication bypass (CVE-2026-76536, 9.4). On Java4Shell, Red Hat’s RHEL errata shipped on 19 November — correcting yesterday’s briefing, which said Red Hat had no fixed build — and Temurin’s Java 17 fix is now out; GreyNoise still reports probing only, with no successful intrusions. The European Commission confirmed on 19 November that it responded to an AI-generated cyber campaign detected on 22 October; a report that ARTEX was used is unconfirmed by the Commission. Singapore’s MAS warned against giving AI agents bank login credentials, and the UK’s Bank of England, FCA and PRA opened a joint cyber-resilience consultation.
Confirmed, reported and unconfirmed items are labelled as such. No MITRE mapping is shown where the source published none.
Global card networks and the largest US bank; premium-cardmember marketing data held by third-party vendors
A financial institution running Temenos Quantum Fabric; six further banks in Asia, the Middle East, Europe and the Americas share the configuration
PayU Payments Private Limited (India); payout API and cloud servers
Australian stockbroking and wealth firm; clients’ data
KPMG Canada clients; tax documents
National Credit Information Center (CIC), the State Bank of Vietnam’s credit-data hub
South Korean savings-bank staff (training-management system run by education provider KG Eduone)
Japan’s largest online broker; up to about 15.72 million accounts
Irish credit unions; saved payee records (IBANs, account numbers, sort codes)
Japanese firms using Plaid’s KARTE customer-experience platform; MUFG Bank, Mitsubishi UFJ Morgan Stanley Securities, Higo Bank and others
European Commission (HR payslip / time-management and medical-reimbursement systems, per press reporting)
US fintech; parent of Square and Cash App
Argentine finance firm
US bank (Comerica Bank)
HSBC Taiwan private-banking and wealth-management clients
Godrej & Boyce, PlasmaGen Biosciences and Connectwell Industries (Indian corporates, not financial firms); EKwiiK / EazyPay (Botswana payments platform)
US bank customers (Chase, Capital One, Citi, Wells Fargo logins found in stolen records)
No new development in-window: SBI (the FIR covers the October incident already recognised as stale); Binance (no new development — 87 networks reopened, full reopening aimed at about 22 November); Coinone / MetaM (the FSS virtual-asset CISO meeting was held on 19 November; no outcome or FSS statement has been reported); Barracuda and Infoblox (no new development); KOMSCO chak (no new development in-window); the Korean FSI special audit (self-assessments still due 30 November, no new status); PANZER vs Bank Jabar Banten Syariah and ERBANK (no bank response and no new reporting in-window; both remain leak-site claims only); Goodbody; Storm-1175 — Insight Credit Union and Sawyer Savings Bank; Bank SinoPac (full report still unconfirmed as filed); Caixa Vida e Previdência; Tesouro Nacional; KB Capital; Flybit; SMBC Nikko Securities; Seven Bank; Mirae Asset Securities; Philippine SEC; Bank Rakyat; INC Ransom — Raymond James Financial, Game Changer Wealth, JETS Jamaica, Shenwan Hongyuan (H.K.) and StrucSure Home Warranty (claims still unverified); AXA Direct Japan and Monex Securities (no new figures); KBIZ; YES Bank denial and Tamil Nadu arrests (covered 15 November, no new development); TodoPago, Endzone and Payoutsking (claims still unverified); Bank of Khartoum (Bankak allegations, still unconfirmed); Octopus Hong Kong (still no corroboration); Hirudo (no new development); FaceStealer (no new facts); Inception / Paychex (no new development); N0n / D.R. Johnson Financial Group (no new development); Maileva (no new development); Deutsche Bank whistleblower files (no new stories drawn from the files in-window; the Hesse data-protection probe opened on 19 November is covered in Supply chain). Absence of reporting is not evidence of absence of activity.
No new CISA KEV additions were made on 19 November — verified against the live feed in this run (catalogue version 2026.11.18, 1,786 entries). The four 18 November additions (Synology CVE-2024-10441, LiteLLM CVE-2026-40217, Apache Axis CVE-2026-90436, REDCap CVE-2026-90817) fall due 21 November, except LiteLLM (2 December), and are carried below. A due date alone is not a new development under the standing rule. Several pages republished on 19 November headline “CISA adds four” for Adobe Commerce, Windows and N-able flaws — those were added on 8 September 2026, not 19 November, and are not reported as new here.
| CVE ID | CVSS score | Affected product | Exploitation status (PoC / in the wild) | Relevance to finance |
|---|---|---|---|---|
| CVE-2026-123890 — NEW, zero-day exploited since September | 9.8 | VMware vCenter Server (Directory Service); vSphere / VCF | In the Wild (confirmed zero-day); no public PoC; not in CISA KEV | Unauthenticated authentication bypass in the VMware Directory Service (SRP-6a session-key disclosure over DCE/RPC, TCP 2012). Google Threat Intelligence Group / Mandiant disclosed on 19 November that suspected PRC-nexus actor ARMORY CASTLE (formerly UNC5221) exploited it as a zero-day in early September 2026, using a custom Go tool (BONESAW) to create a vSphere single sign-on administrator account, exfiltrate the directory database and steal the STS SAML signing key to forge admin assertions; a second tool (SAWSNAKE) was deployed after. The group reached vCenter after pivoting from a BeyondTrust Remote Support zero-day (CVE-2026-102764). Broadcom disclosed and patched the flaw the same day in advisory VMSA-2026-0009: fixed in vCenter 9.1.1.0200, 9.0.2.0200 and 8.0 U3m, and VCF 5.x 5.2.5; vCenter 7.0 only via extended support. There is no workaround; Broadcom calls it an emergency change. The same advisory fixes four more flaws (CVSS 6.7–8.1 range): SDDC Manager SQL injection CVE-2026-123891 (8.1), command injection CVE-2026-123892, privilege escalation CVE-2026-123893 and VMware Tools SAML token bypass CVE-2026-123894. No indicators of compromise were published in the sources read. vCenter is the control plane for bank virtualisation; a stolen SAML signing key forges administrator assertions across the estate. Patch now and hunt for unauthorised SSO administrator accounts created since early September. |
| PHP November security release — NEW, seven critical RCE flaws | 9.2 (CVSS 4.0, PHP project) | PHP 8.2, 8.3, 8.4, 8.5 (and 8.6.0) | No exploitation reported; no verified public PoC | PHP released security updates on 19 November (8.2.36, 8.3.37, 8.4.28, 8.5.13 and 8.6.0), fixing 17 CVEs, seven of them rated critical (9.2 CVSS 4.0): CVE-2026-115560 (session serializer mismatch leading to remote code execution), CVE-2026-126714 (“phpAfterShell” — TOCTOU partially-overwritten session files leading to RCE, needing attacker-influenced session data, an interrupted write and a gadget chain), CVE-2026-129289 (PHP-FPM request smuggling leading to unauthenticated RCE), CVE-2026-129496 (timezone cache use-after-free leading to RCE), CVE-2026-129481, CVE-2026-129474 and CVE-2026-129182. CVE-2026-115560 was pre-announced on 23 October with the mitigation session.upload_progress.enabled=0; the PHP Foundation says the combination it needs is very unlikely in practice. The release also fixes flaws in PHP-FPM, PHP-CGI, the timezone cache, PHAR, SOAP and the HTTP stream wrapper. No CVSS scores have been published for individual CVEs in the batch beyond the project’s own ratings, and no exploitation of these flaws has been reported. PHP 8.1 and older are no longer supported and receive no fix. PHP underpins bank web applications and payment pages; the session-handling flaws are the priority, and the FPM smuggling flaw is unauthenticated. |
| CVE-2026-76536 / CVE-2026-76535 — NEW, Splunk November update | 9.4 / 9.1 | Splunk Enterprise (standalone and clustered instances) | Not exploited; no public PoC; not in CISA KEV | Splunk’s November 2026 update fixes 146 flaws; the CVE records were published on 19 November (20:17 UTC / 20 Nov 04:17 SGT). Headline flaws: CVE-2026-76536 (9.4) — unauthenticated authentication bypass via an empty clustering shared secret on standalone instances, allowing an attacker to read or modify cluster configuration via the REST API; and CVE-2026-76535 (9.1) — unauthenticated SPL execution as a valid user via RapidDiag. A further remote code execution flaw, CVE-2026-131253, is in the batch; its CVSS was not verified in this run. Fixed in Splunk Enterprise 10.6.1 / 10.4.5 / 10.2.9 / 10.0.12 / 9.4.17. Workaround for CVE-2026-76536: set a strong unique pass4SymmKey (at least 12 characters) in server.conf. Splunk is the SIEM at the centre of bank SOC detection; an unauthenticated bypass on it blinds the defenders first. |
| CVE-2026-132165 — NEW, brief note (advisory 18 Nov, reported 19 Nov) | — no CVSS yet (Drupal risk 18/25, Critical) | Drupal <10.6.22, 11.0.0–11.3.21, 11.4.0–11.4.12 | PoC exists per advisory; no in-the-wild exploitation reported; not in CISA KEV | Critical installer remote code execution: an unauthenticated attacker can use install.php to attempt reinstalling a site, leading to takeover and remote code execution. Fixed in 10.6.22 / 11.3.22 / 11.4.13 (released 18 Nov). Found by Anthropic’s OSS Scanner. A companion flaw, CVE-2026-132166, is a moderately critical gadget chain needing admin access. Bank marketing and content sites running Drupal should patch, and check whether install.php is reachable. |
| CVE-2026-123007 (“NGINX2Root”) — NEW | 9.2 / 8.1* | NGINX Open Source 1.9.5–1.31.8 (HTTP/2 enabled); NGINX Plus | Public PoC; no in-the-wild exploitation reported; not in CISA KEV | Integer underflow (CWE-191) in HTTP/2 flow-control accounting, present since NGINX 1.9.5 (2015): an unauthenticated attacker can achieve remote code execution in default configurations, including static-file servers with no backend. Rapid7 (Stephen Fewer) published a working proof of concept on GitHub on disclosure day, demonstrating unauthenticated RCE against nginx 1.31.8 in Docker (linux/amd64, TLS + HTTP/2). Limits: it needs about 128 GB of HTTP/2 traffic on one connection, a specific x86-64 glibc memcpy path (AVX-512), and likely fails through a proxied HTTP/2 connection. F5 says code execution requires ASLR disabled or bypassed; Wiz reports bypassing ASLR, and the discoverer Xint chained the HTTP/2 bug with an unpatched Linux kernel SCTP flaw to reach root, including from a Docker container. Wiz and Xint are withholding their exploits. Warning: fake nginx2root exploits are circulating on GitHub — do not run exploit code from untrusted repositories. Fixed in NGINX Open Source 1.30.7 (stable) / 1.31.9 (mainline) and NGINX Plus 37.0.8.1 / 37.3.0.1. New in-window (reported): Threatbear firmware analysis found Ubiquiti’s current UniFi Cloud Gateway Ultra firmware still ships a vulnerable nginx with HTTP/2 enabled on four ports; no Ubiquiti advisory has been published yet. No in-the-wild exploitation has been reported as of 19 November, and none of the five flaws is in CISA KEV. Workaround if patching must wait: temporarily disable HTTP/2. NGINX sits in front of banking web, API and payment-gateway services; Wiz puts its footprint at 225 million-plus websites. *F5 rates it 9.2 (CVSS 4.0) / 8.1 (CVSS 3.1). |
| CVE-2026-123633 — NEW | 9.2 / 8.1* | NGINX Open Source 1.19.7–1.31.8 (HTTP/2 as proxy); NGINX Plus | No public PoC; no in-the-wild exploitation reported; not in CISA KEV | HTTP/2 connection use-after-free; Wiz demonstrated remote code execution with an ASLR bypass. No public proof of concept has been released by Wiz. Fixed in NGINX Open Source 1.30.7 / 1.31.9 and NGINX Plus 37.0.7.1 / 37.2.0.1. No in-the-wild exploitation has been reported as of 19 November; not in CISA KEV. Three further critical flaws were disclosed in the same F5 batch, all rated 9.2 CVSS 4.0 by F5: CVE-2026-123634, CVE-2026-121810 and CVE-2026-113425 (resolver, HTTP/3 and mail-proxy paths). All five flaws were found independently by Wiz (Atlas AI harness), Ikotas Labs, Metnir, Rapid7 and Xint. *F5 rates it 9.2 (CVSS 4.0) / 8.1 (CVSS 3.1). |
| CVE-2026-129448 — NEW, catch-up (missed by the 18 Nov edition) | 9.8 | Apache Roller 4.0.0–6.1.4 (same bug as CVE-2026-82384 for 6.1.5) | In the Wild (Apache confirms active exploitation) | Deserialization of untrusted data in the XML-RPC endpoint: the endpoint accepts vendor extension types and deserializes them before authenticating the requester, letting an unauthenticated remote attacker execute arbitrary code. Apache’s notice states plainly: “This issue is being actively exploited.” Apache has not said who is behind the attacks or how many servers were hit, and has disclosed no technical details about the attacks. The advisory was published on 17 November and widely reported on 18 November, but was missed by the 18 November edition. Fix: upgrade to Roller 6.1.6 or later, which disables the extension types and rejects XML-RPC requests when the XML-RPC feature is disabled; users who cannot upgrade at once should set enabledForExtensions to false — disabling XML-RPC in the administration UI alone is not enough, because older versions deserialize the request before checking that setting. Researcher n0mi1k is credited with discovering the flaw; Matt Raible reported it. Roller is a niche Java blogging server, but any instance reachable on a bank’s domains is an unauthenticated RCE path. |
| CVE-2024-10441 — NEW, added to CISA KEV 18 Nov | 9.8 | Synology DSM / BeeStation OS (system plugin daemon) | In the Wild (KEV); no PoC identified | Improper output encoding (CWE-116) in the system plugin daemon, allowing unauthenticated remote code execution. CISA added it to the Known Exploited Vulnerabilities catalogue on 18 November (federal due date 21 November, forensic triage required). A fix has been available since December 2024: DSM 7.2.2-72806-1, 7.2.1-69057-6, 7.2-64570-4, 7.1.1-42962-7, 6.2.4-25556-8; BeeStation OS 1.1-65374. Context: Synology said on 29 October that ransomware is encrypting outdated DSM devices; which flaw is being used is unresolved (Synology points to a December 2024 flaw without naming it; firm 360 attributes cases to CVE-2026-13639 / CVE-2026-13684 and C3RB3R / Cerber). NAS devices sit in small offices and branches and are often unpatched — treat any unpatched Synology device as potentially compromised and check it before patching. |
| CVE-2026-90817 — NEW, added to CISA KEV 18 Nov | 9.8 | Vanderbilt REDCap (research data capture) | In the Wild (KEV); public PoC (Nuclei template) | Code injection (CWE-94 / CWE-73), unauthenticated with a valid public survey link; found by Securifera. CISA added it to KEV on 18 November (federal due date 21 November, forensic triage required). Fixed in 16.0.49, 17.3.10 and 17.4.4. BFSI relevance is low — REDCap is mainly used by universities, hospitals and research organisations — and it is included here as a KEV item only. |
| F5 BIG-IP — six critical flaws — NEW | 9.8 / 9.3* | F5 BIG-IP 21.1.0, 17.5.0–17.5.1, 17.1.0–17.1.3 (Traffic Management Microkernel) | No exploitation reported; no workaround | Stack-based buffer overflows (CWE-121) in the Traffic Management Microkernel (data plane): CVE-2026-123637, CVE-2026-121805, CVE-2026-120292, CVE-2026-120285, CVE-2026-116953 and CVE-2026-120722. Each causes an unauthenticated TMM crash (denial of service until restart); F5 says remote code execution is possible. Each flaw needs a specific virtual-server setup (an iRule using PEM::subscriber, MR::message route, PSC::ip_address or GENERICMESSAGE::route, DNS::scrape, or an APM per-request policy with Client Information plus Endpoint Security agents). No workaround exists — fixes are engineering hotfixes only, which F5 notes get less QA than scheduled releases, so test before production. WatchTowr’s Sina Kheirkhah reported four of the six. BIG-IP fronts banking applications and APIs. *F5 rates them 9.8 (CVSS 3.1) / 9.3 (CVSS 4.0). |
| CVE-2026-114938 (“Java4Shell”) — STATUS, Red Hat fix lands 19 Nov — correction to the 19 Nov edition | 8.1 / 9.2 / 9.8* | Oracle Java / OpenJDK, HotSpot JVM on x86-64 with AVX-512 VBMI; OpenJDK 17.0.6 and later (also 21, 25, 27) | No confirmed exploitation; no verified public PoC repository; not in CISA KEV | Correction — yesterday’s briefing said Red Hat had no fixed build; that is now wrong. Red Hat’s Build of OpenJDK fixed builds — 17.0.21.1, 21.0.13.1 and 25.0.5.1 — shipped on 18 November (RHSA-2026:100565, 100566, 100569, 100570 and 100573), and RHEL 8, 9 and 10 errata followed on 19 November (RHSA-2026:100563 and 100564 for Java 17; 100567 and 100568 for Java 21; 100571 and 100572 for Java 25). UBI OpenJDK container images have not yet been rebuilt. Temurin’s Java 17 fix is now available (17.0.21.1+2); Temurin Docker images are still not rebuilt as of 19 November. Keycloak 26.8 container images now ship fixed OpenJDK 21.0.13.1. GreyNoise reports 192 Java4Shell-shaped requests from 6 IPs as of 19 November, with zero successful intrusions; WatchTowr honeypots are seeing java4shell-recon user-agents plus blind exploit attempts. Memory corruption / out-of-bounds write in the HotSpot JVM’s AVX-512 Base64 decoder: specially crafted Base64 input can crash the JVM and, in the worst case, allow unauthenticated remote code execution. CSA Singapore issued alert AL-2026-166 on 18 November and CERT-EU published advisory 2026-021 the same day. Siemens published advisory SSA-272816 on 18 November: Industrial Edge Management Pro V1 and V2 and Industrial Edge Management Virtual are affected in all versions with no fix available — block direct internet access. Wiz proved remote exploitation feasible against Jenkins, Confluence and Apache Tomcat — not Jira. No new verified public PoC repository was found — the cve.now count rose from 24 to 55, none of them verified, and none is linked. CISA’s alert (17 Nov, joint with Canada’s Cyber Centre) still urges immediate patching, starting with internet-facing systems; Oracle released fixed builds on 17 November as part of its November 2026 CPU. Workaround: the -XX:UseAVX=2 flag disables the AVX-512 path. Java underpins core banking, payments and identity stacks, and CISA warns bundled runtimes in containers and appliances may not appear in software inventories. *Oracle rates the CVE 8.1 (CVSS 3.1); researcher Calif rates it 9.2 (CVSS 4.0); Red Hat’s preliminary 9.8 has no CVSS 4.0 vector published. Do not merge the scores. |
| CVE-2026-115731 — NEW, added to CISA KEV 17 Nov | 9.4 | Perforce Puppet Enterprise / Puppet Server (report processing) | In the Wild (KEV); public PoC | Deserialization of untrusted data (CWE-502) in Puppet Server report processing: any client holding a CA-signed agent certificate can submit a crafted report via PUT and execute arbitrary code on the Puppet Server as its service user, and code execution as the puppet user gives access to the CA signing key. CISA added it to the Known Exploited Vulnerabilities catalogue on 17 November, which means CISA has evidence of exploitation; the federal due date is 1 December. Fixed in Puppet Enterprise 2023.8.13, 2025.12.1 and 2026.0.1 (released 3 November), Puppet Core 9.1.3 / 8.22.1 and OpenVox Server 9.2.0 / 8.18.0. Neodyme / OpenVox published a working proof of concept with the 3 November fix. Open-source Puppet has no patched release, per Vox Pupuli reporting. A companion flaw, CVE-2026-115732 (CVSS 4.0 9.0, certificate-name authorisation bypass), was fixed the same day but was not added to KEV on 17 November. A compromised Puppet Server controls configuration on every managed host. CISA’s required action includes checking whether a system was compromised before patching, under BOD 26-04. |
| CVE-2026-119618 — STATUS, CSA AL-2026-164: reportedly exploited | 8.1 / 9.8* | Apache HTTP Server 2.4.55 to before 2.4.71 with mod_http2 enabled (standalone mod_h2 2.0.8–2.0.42) | Reportedly exploited (CSA); public PoC | CSA Singapore issued alert AL-2026-164 on 17 November: the race condition (CWE-362) in assign_headers() in Apache HTTP Server’s mod_http2 module is “reportedly being actively exploited, and a proof-of-concept exploit is publicly available”. An unauthenticated attacker can achieve remote code execution on servers with HTTP/2 (h2/h2c) enabled. Fixed in Apache HTTP Server 2.4.71, released 11 November; disabling HTTP/2 is a workaround. The 12 November research proof of concept reproduces heap corruption and crashes on 2.4.70 — it does not demonstrate code execution — and Sysdig (13 Nov) and Abijita (12 Nov) both stressed there was no public RCE proof of concept and no confirmed in-the-wild exploitation as of 12–13 November. CSA does not say who reported the attacks, where they were seen, or how many servers were hit, and the flaw is not in CISA KEV as of 17 November (VulnCheck’s KEV feed listed it as exploited in the wild on 15 November; that is VulnCheck’s feed, not CISA KEV). Apache HTTP Server is widespread in web-facing banking infrastructure. *CSA scores it 8.1 (CVSS v3.1); Red Hat scores it 9.8; Apache itself published no CVSS. Do not merge the scores. |
| CVE-2026-130796 — NEW | 8.2 | Flowise 3.1.3–3.1.4 (AI application builder) | No exploitation known; no fix — project archived | MCP server bearer-token disclosure (CWE-212): Flowise returns an MCP server bearer token in plain text via GET /api/v1/public-chatflows/:id. Anyone reading a public chatflow can then call POST /api/v1/mcp/:chatflowId, bypassing API-key protection and consuming the owner’s model quota. Published 17 November (VulnCheck advisory; researcher George Chen reported the flaw privately on 9 June 2026). No fixed version exists — Flowise was archived on 10 August 2026 and core support ended 31 August 2026, so exposed self-hosted instances stay exposed. Any bank or fintech running self-hosted Flowise must isolate or retire it, not wait for a patch; rotate any exposed MCP token. |
| CVE-2026-130615 — NEW | 7.1 | Flowise through 3.1.4 (AI application builder) | No exploitation known; no fix — project archived | Server-side request forgery via DNS rebinding (CWE-367) in the Playwright / Puppeteer loaders, allowing an authenticated user to reach internal and cloud-metadata endpoints. Published 17 November (VulnCheck advisory). As with CVE-2026-130796, no fixed version exists — the project is archived — so exposed self-hosted instances stay exposed. |
| CVE-2026-106173 — STATUS, PoC verified genuine | 9.4 | Fortinet FortiSIEM | Public PoC (verified genuine) — in-the-wild exploitation contested | New in-window development: the public proof of concept is verified genuine. It is Horizon3.ai’s exploit (20 October, tagged as the exploit reference in the CVE record): Python, over gRPC on TCP 27918, overwriting a cron-run script, with execution delayed about 60 seconds. A critical path-traversal flaw in FortiSIEM, disclosed in October and patched by Fortinet, that could allow unauthenticated attackers to execute arbitrary code or commands. The exploitation picture remains contested: Canada’s cyber centre said open-source reporting indicates in-the-wild exploitation (23 October); Fortinet says it is not aware of exploitation; the flaw is not in CISA KEV. Confirm the FortiSIEM patch is applied. No new in-window development beyond the PoC verification. |
| CVE-2026-84869 — NEW, exploitation confirmed | 9.9 | ConnectWise ScreenConnect, technician host client below 26.6.5 | Exploitation confirmed (Huntress, 12 Nov incident) | A guest in a ScreenConnect session triggers file transfers the technician’s host client runs without authorisation. Huntress confirms a full technician-workstation compromise in about nine seconds on 12 Nov and more than 688,000 guest “run file” requests across 42 MSP servers from 20 Oct to 16 Nov. In CISA KEV since 11 September. Fix: upgrade technician clients to 26.6.5+ — the server patch alone does not fix it — or remove the TransferFiles permission. A technician compromise can cascade to every client the MSP serves, including financial firms. |
| CVE-2026-116606 / CVE-2026-111560 — NEW alert (CSA AL-2026-162, 16 Nov) | 9.8 / 7.2 (CSA v3.1) vs 9.3 / 8.6 (Barracuda / watchTowr v4) | Barracuda Email Security Gateway, Load Balancer ADC, Message Archiver, WAF Control Center, WAF, Web Security Gateway | In the Wild (KEV since 10 Nov); public PoC | Actively exploited, chained: authentication bypass (CVE-2026-116606) with OS command injection (CVE-2026-111560) for unauthenticated root on an exposed Email Security Gateway admin interface. CSA Singapore issued alert AL-2026-162 on 16 November. The CVEs were covered in the 14 November edition; the CSA alert was covered in the 17 November edition. Do not merge the two scoring versions. A proof of concept was published by watchTowr Labs about 3 November, six days before Barracuda confirmed exploitation on 9 November against “a small number of organizations” — the repository URL was not verified first-hand and is not linked. Fix: ESG 9.4.0.030; CSA says update to the latest versions immediately. Email gateways sit in front of bank mail flow; unauthenticated root on one is a full-compromise path. |
| CVE-2026-99288 (“NoisyTCP”) — NEW | 7.5 (Red Hat draft) / 8.1 (Oracle) / Medium (Canonical); NVD unscored | Linux kernel TCP, 4.4+ (race condition in the lockless TCP established-socket lookup, latent since October 2015) | No public exploit — remote kernel panic demonstrated | Unauthenticated remote kernel panic (denial of service) via crafted TCP packets to any listening port, including Nginx, Apache and SSH defaults. No remote code execution was demonstrated. Full write-up published by researcher Hyunwoo Kim on oss-security on 15 November 16:00 UTC (16 Nov 00:00 SGT). Fixed kernels 6.12.117, 6.18.63, 7.3.4 (plus a second patch for arm64 / POWER / RISC-V); Debian, Oracle UEK and CIQ Rocky have shipped fixes, while RHEL 8/9/10 and Ubuntu had not as of 16 November. Not in CISA KEV. Linux servers and cloud infrastructure underpin bank systems; a remote unauthenticated panic against internet-facing hosts is a live availability risk. Do not merge the conflicting scores. |
| CVE-2026-129349 — NEW, catch-up (missed by the 16 Nov edition) | 9.1 | Apache Shiro through 3.0.0 (DefaultLdapRealm / JndiLdapRealm) | No confirmed exploitation; no verified public PoC | LDAP authentication bypass: with authenticationCachingEnabled=true (not the default), a user who logged in once can then log in with any password. Disclosed 14–15 November in a batch of 17 Apache flaws; fixed in Shiro 3.0.1. Aggregators list a GitHub repository (googlesky/CVE-2026-129349) as a public PoC, but it was not verified as genuine and is not linked. Shiro is a Java authentication framework used in enterprise applications and APIs. The non-default precondition limits exposure, but any deployment with LDAP caching enabled should patch. |
| CVE-2026-126259 — NEW, scope expansion 16 Nov | 9.0 | FortiOS 8.0.0, 7.6.0–7.6.7, 7.4.0–7.4.12, 7.2.0–7.2.13, and 7.0 all versions (added 16 Nov) | Not known exploited | Missing authentication (CWE-306): where HA FGSP with standalone-config-sync or autoscale is in use, an unauthenticated attacker can execute commands as super admin via crafted HA packets. Fortinet PSIRT FG-IR-26-273, first published 10–11 November; the 16 November update added FortiOS 7.0 all versions to the affected list. Fixed in 8.0.1 / 7.6.8 / 7.4.13 / 7.2.14; virtual patch FG-VD-61445.0day is available. FortiGate is edge infrastructure for banks; an unauthenticated super-admin path on HA pairs is a patching priority where the configuration applies. |
| 4EverAble2Shell / Agent2Shell — CVE-2026-90436 + VULN-2026-0009 — STATUS, HF3 shipped 16 Nov | 9.8 (CVE-2026-90436) | N-able N-central (on-premises servers with Legacy API enabled); Atlassian Confluence Data Center 1.x–9.x with Remote API enabled | In the Wild (KEV, added 18 Nov) | Hotfix 3 released 16 November — phased rollout for hosted environments began that day, on-premises general availability is open, and the Legacy API is now permanently blocked. Until HF3 is applied, Hotfix 2 plus Legacy API disabled plus Agent / Probe upgrades remain the guidance. Rapid7 confirms CVE-2026-90436 exploitation against Confluence Data Center on 13 November and against N-central on 13 November; a PoC targeting N-central was published 12 November. Huntback counted 21 production N-central servers mining XMRig, 19 of them on the Hotfix 2 build. Apache Axis 1.x is end-of-life with no fixed release planned; Confluence is fixed in 9.2.28 and 10.x is unaffected. PoC repo: MSPNightmare/CVEs — N-Central N-Able (GitHub) — genuine researcher disclosure (Sean Bales), with Previdian’s caveat that the public PoC is unverified and may contain malware. CISA added it to KEV on 18 November (federal due date 21 November, forensic triage required). RMM used by MSPs serving small banks; upgrade to HF3 now and hunt the IOCs in Indicators. |
| CVE-2026-121653 — STATUS, PoC status contested | 9.8 | Ivanti Endpoint Manager Core, 2024 SU7 and earlier | Public exploit reported — contested; in-the-wild exploitation not confirmed | No new development in-window. Unauthenticated SQL injection leading to remote code execution on the endpoint-management server. Fix: EPM 2024 SU8; no workaround published. Ivanti notes it is not externally exploitable in dual-NIC deployments and says it is not aware of any customers being exploited. The public-exploit picture is now contradictory: SANS (12 Nov) said no public working exploit, while a Week-46 roundup lists it as a public exploit with no verifiable repository — no genuine PoC repository was found, so no repo is linked and the PoC should not be presented as confirmed. VulnCheck added the CVE to its KEV feed on 15 Nov — that is VulnCheck’s feed, not CISA KEV; do not present it as a CISA listing. Not in CISA KEV. Sibling critical in the same 10 Nov batch: CVE-2026-121683 (9.4, missing authentication in Cloud Services Appliance 5.0.8 only; fix 5.0.8.1). EPM Core is the management server for bank endpoints; unauthenticated RCE on it is a full-compromise path. |
| CVE-2026-76466 — NEW, catch-up | 9.8 | Cisco Expressway, clustered Expressway-C / Expressway-E, X15.5.3 and earlier | Public PoC — exploitation not confirmed | Catch-up — Cisco advisory 21 Oct; PoC repository created 5 Nov; press coverage 12 Nov; never covered in a prior edition. Hard-coded credentials (CWE-798): a TLS 1.3 connection to TCP/4372, an Erlang distribution handshake as wombat@127.0.0.1 with the hard-coded cookie wombat, then rpc:call(os, cmd) as root. Fix: X15.5.20; X14 and earlier must migrate; no workaround. Validated only on Expressway-C X15.3.2, a two-node cluster in TLS Permissive mode (the factory default); TLS Enforcing mode is untested. Not in CISA KEV; Cisco PSIRT is aware of no malicious use. PoC: GitHub dinosn/cve-2026-76466 (Nicolas Krassas, created 5 Nov; exploit, fleet detector and docs — verified genuine). Do not link the Feedly-listed copy emirwtz/CVE-2026-76466 — not verified. Expressway is video-conferencing edge in banks; restrict TCP/4371–4372 to cluster peers until patched. |
| CVE-2026-126173 — NEW | 9.8 | AutomatorWP Pro Webhooks add-on ≤ 1.4.4 | Public PoC — in-the-wild exploitation not confirmed | Missing authorisation in the webhook handler lets an unauthenticated remote attacker reset any user’s password, including administrators. Published 14 Nov (Wordfence CNA); public PoC published the same day. PoC: GitHub im0xfay/CVE-2026-126173 (created 14 Nov; contains poc.py, a Docker lab and analysis — verified genuine, but it is a third-party reproducer, not the original discoverer’s repo; the README is in Indonesian and it includes mass-scan and mass-exploit modes plus a claimed live run against two FOFA targets on 14 Nov — the mass-exploit tooling is the risk escalator). No fixed version was listed at publication. PatchCVE’s claim of “public exploitation activity reported” is an unverified aggregator assertion — treat as unconfirmed. Not in CISA KEV. Bank marketing and content WordPress sites running this add-on face admin takeover, defacement and customer phishing. |
| CVE-2026-19907 — NEW | 9.8 | Shipo WordPress plugin 1.6–1.8.0 | No PoC reported | Unauthenticated admin authentication bypass: authenticate_rest_user() trusts a username taken from the LOGGED_IN_COOKIE without validating its signature, so knowing a username (for example admin) is enough. Disclosed 14 Nov (Wordfence CNA). No PoC reported; not in CISA KEV. The WordPress directory offers 1.8.4, but its changelog does not name this CVE — the fixed version is not independently confirmed. Same WordPress-site caveat as the AutomatorWP flaw, but no PoC and a narrower plugin. |
| CVE-2026-124969 — NEW, brief note | 7.1 | Optimole WordPress plugin ≤ 4.2.19 | No PoC reported | Unauthenticated arbitrary file deletion (via comment_content / comment_agent during attachment rename) that can lead to remote code execution by deleting wp-config.php; it needs PEAR Archive_Tar present and a subsequent admin rename action. Published 14 Nov. No PoC; not in CISA KEV. |
| CVE-2026-20320 — STATUS, vendor walk-back | — not stated in sources read | Cisco BroadWorks | In the Wild (KEV) — vendor evidence softened | Cisco advisory v1.2, dated 13 Nov, corrected the record: Cisco states it does not have direct evidence of active exploitation, walking back the 12 Nov update that reflected known active exploitation. The CISA KEV listing (added 10 Nov) stands. Treat the exploitation picture as contested between the KEV listing and the vendor’s own evidence position. |
| CVE-2026-109618 / 109619 (PHP2Shell) — STATUS, Siemens downstream notices | 9.2 | PHP; PAN-OS and other network/security products; Check Point Security Gateways | In the Wild | New in-window development: Siemens confirmed PHP2Shell affects RUGGEDCOM APE1808 running PAN-OS and several building-automation and energy products (SSA-566512; CISA ICSA-26-316-05) — industrial, not BFSI, so one line only. The Lazarus targeting discussion (Cyperma via Yonhap, 15 Nov 08:43 SGT) was covered in the 15 November edition; the attribution is Cyperma’s, reported not independently confirmed. Check Point’s detail adds that the TSGA/TSProxy implants are linked to the PhiliKit toolset (which ESET ties to UNC5221), but Check Point says it “cannot assess the operation with high confidence”. GreyNoise telemetry reported in the 13 Nov afternoon deep-research check shows PHP2Shell exploitation continued through 12 Nov. The Check Point campaign detail is unchanged: attackers chained PHP2Shell in the Mobile Access Portal with CVE-2026-123039 to gain root on unpatched gateways, planted the TSGA backdoor, pivoted to a VMware ESXi host and ran a wiper — “a couple of dozen” systems in one region. Both fixes are in the November Jumbo Hotfix. The PHP2Shell fix had been available since 23 Oct — the one-week patch lag is the lesson. Keep management interfaces off the internet; apply the November Jumbo Hotfix. IOCs: 130.94.19[.]204, 130.94.19[.]168; upload_progress_pwn string in PHP session files; process masquerading as CPconfigdaemon; DREAM_ANNEXATION.txt notes. |
| CVE-2026-121529 (“Fortinet2Shell”) — STATUS, PoC count up | Critical (no CVSS figure in sources read) | FortiClientMac 7.4.0–7.4.7; FortiClientWindows 7.4.0–7.4.7 and 7.2.0–7.2.14 | Public exploit — in-the-wild exploitation not reported | New in-window detail: the public PoC count is now five GitHub repositories per cve.now (the 13 November edition listed two). Still no confirmed in-the-wild exploitation; not in CISA KEV. Origin-validation flaw, unauthenticated remote code execution on FortiClient endpoints. PSIRT FG-IR-26-194. FortiClient is the endpoint agent on bank staff machines; an unauthenticated RCE with a public exploit is a patching priority. |
| CVE-2026-6460 — NEW, catch-up | 8.8 | PostgreSQL 14–18 (flaw dates to 2004) | Public partial PoC — in-the-wild exploitation not reported | Catch-up — patch published 12 Nov, researcher write-up 13 Nov; missed by the 14 November edition. A type-confusion flaw in PostgreSQL core (composite types plus ALTER TYPE between PREPARE and EXECUTE) lets an authenticated database user with object-creation privileges execute arbitrary code as the OS user running the database server. CVSS 8.8. It was one of 52 flaws fixed in the 12 Nov release (32 scored 8.8); the researcher traces the primitive to May 2004 — effectively every PostgreSQL since then. Researcher Mehmet Ince published the crash PoC and address-leak steps on 13 Nov and deliberately withheld the final RCE step, so there is a public partial PoC, not a copy-paste weapon. Full protection also requires rebuilding C-language extensions. PostgreSQL 14 is now end-of-life (14.25 is its last release). Sibling flaw CVE-2026-116322 lets an unauthenticated client trigger a restart of PostgreSQL 18 servers (denial of service). Not in CISA KEV. PostgreSQL underpins many banking and fintech stacks; authenticated RCE as the database OS user is a high-priority patch. |
| CVE-2026-124194 — STATUS, Patch Tuesday follow-up | 6.5 | Microsoft Outlook (NTLMv2 hash disclosure) | In the Wild (KEV) | No new development in-window. November Patch Tuesday is the largest on record by Qualys’s count — 1,252 CVEs this month (Qualys, 13 Nov). Conflict — flag: Microsoft’s own count differs (1,362 CVEs across 27 updates per other trackers) and Tenable counted 1,093 — a counting-method difference between trackers, not a new item. About 35% of November patches carried a Low reliability rating in Qualys’s assessment as of 11 Nov. Outlook CVE-2026-124194 remains the exploited item in the batch. Microsoft confirmed the Windows 11 update KB5132061 fails to install on some devices that took the October non-security preview KB5127936, and KB5131085 (Windows 10 / Server 2019) is reported to break Exchange — both affect deployment planning. |
| CVE-2026-127063 — carried | 9.3 | DataComm RMM+ extension for ConnectWise ScreenConnect | Not known exploited — no PoC reported | No change in-window. “The third-party RMM+ extension for ConnectWise ScreenConnect uses a static API secret by default.” Caveat: the CVE wording (“static API secret”) differs from Huntress’s account of a blank access key accepted when none was configured — treat as the same extension weakness pending vendor confirmation. This is the extension weakness behind the SCKit worm. Check RMM+ versions 1.0.31/1.0.32 and remove or re-key the extension. |
| Able2Shell — seven CVEs — carried | 9.3 (highest) | N-able N-central | PoC public — exploitation of new chains claimed, contested | No change in-window except the 4EverAble2Shell row above. Original Able2Shell (9 Nov): researcher Sean Bales published PoC code for 33 alleged N-central flaws. Hotfix N-central 2026.4 HF2 (build 2026.4.1.26), released 10 Nov, fixes seven CVEs: CVE-2026-125883 (XXE, 9.3), CVE-2026-125878 (OGNL injection), CVE-2026-125879, CVE-2026-125880, CVE-2026-125881 (SSRF), CVE-2026-125882 (RCE via path traversal), CVE-2026-91000. All hosted servers upgraded by 11 Nov 01:04 UTC. The Able2Shell PoC repository (Oracle-Security/CVEs, path “0-Days/N-Central N-Able”, 26 Python PoC scripts) is a genuine researcher disclosure — linkable as the PoC source, with the caveat that N-able has not validated all 33 findings. |
| CVE-2026-123864 — carried | High (no CVSS published) | Node.js 22.x / 24.x / 26.x (HTTPS servers) | Not known exploited — no PoC reported | No change in-window. HTTPS heap-memory disclosure: a zero-length TLS write re-enters the HTTP parser, and the server copies other connections’ heap memory — including TLS private keys and other users’ requests — into a request body. Fixed in 26.13.1 / 24.22.3 / 22.23.5. Node.js published no CVSS scores and lists no workarounds — upgrade only. |
| Node.js batch — six further High — carried | High (no CVSS published) | Node.js 22.x / 24.x / 26.x | Not known exploited — no PoC reported | No change in-window. CVE-2026-123863, CVE-2026-123865, CVE-2026-123866, CVE-2026-123872, CVE-2026-123874, CVE-2026-126051 (HTTP/2 and TLS memory disclosure / DoS / use-after-free). Six Medium and three Low complete the 16-CVE release. Node.js states 16 CVEs; one aggregator lists 17 IDs — report Node.js’s count. |
| CVE-2026-126263 — carried | 9.8 | FortiMail 8.0.0–8.0.1, 7.6.0–7.6.4, 7.4.0–7.4.6 | Not known exploited — no PoC reported | No change in-window. SQL injection (CWE-89): an authenticated attacker can execute unauthorised code/commands via crafted requests. Fixed in 8.0.2 / 7.6.7 / 7.4.7; FortiMail Cloud already remediated. Internally discovered with the Frontier AI model Claude Mythos. |
| CVE-2026-126155 — carried | 9.1 | FortiMail 8.0.0–8.0.1, 7.6.0–7.6.6, 7.4.0–7.4.8, 7.2 all versions | Not known exploited — no PoC reported | No change in-window. Incorrect authorisation (CWE-863): an unauthenticated remote attacker can obtain an authenticated IBE webmail session as any IBE recipient and read or send their encrypted mail. Fixed in 8.0.2 / 7.6.7 / 7.4.9; 7.2 must migrate to a fixed release; FortiMail Cloud remediated 5 Oct. |
| VSA-2026-062 (no CVE ID yet) — carried | — no CVSS assigned | Xen Orchestra / Xen Orchestra Appliance | Actively exploited | No change in-window. Empty-object login token gives Admin privilege; full XCP-ng host takeover as root. Fixed in Xen Orchestra 6.9.1 / 6.10.0. IOCs: 91.211.251.112:8001 and Tor domain ustar3rqxgnskzgnwkgpofgtvxevm6cx3lxdbwt4mndtveqyaohid.onion:3333. |
| SAP November Patch Day — four CVSS 10.0 — carried | 10.0 | SAP NetWeaver AS ABAP / Web Dispatcher / Gateway | Not known exploited | No change in-window. CVE-2026-76964 (ICM memory corruption, unauthenticated RCE — “ICM2SHELL”), CVE-2026-89126 (Web Dispatcher memory corruption, unauthenticated RCE), CVE-2026-89103 (Gateway broken access control), CVE-2026-120017 (SAML2 missing authentication). Patch internet-facing systems first. |
| Fortinet batch — eight criticals — STATUS, public PoC for CVE-2026-126170 | 9.8–9.0 | FortiOS / FortiManager / FortiAnalyzer / FortiSASE | Public PoC — exploitation not confirmed | New in-window development: a public proof of concept for the worst flaw, CVE-2026-126170 (9.8, JSON-RPC API, unauthenticated remote code execution on FortiManager/FortiAnalyzer), has been verified on GitHub (Ghost-Bx1/cve-2026-126170-fortimanager-jsonrpc-rce, created 11 Nov, contains working exploit.py and poc.py — genuine, not a lure); the NVD record was modified 14 Nov. Conflict — flag: PatchCVE states public exploitation activity has been reported, while Fortinet’s advisory says exploitation was not known at publication and cve.now says “public exploit code, no confirmed attacks”; the CVE is not in CISA KEV. Do not present exploitation as confirmed. The JSON-RPC API is disabled by default; the workaround is set rpc-permit none. Fixed in FortiOS 7.2.14/7.4.13/7.6.8/8.0.2 and FortiManager/FortiAnalyzer 7.2.13/7.4.12/7.6.8/8.0.2 — some fixes still “upcoming”; apply the FortiGuard virtual patch, keep OFTP/FGFM ports off the internet. |
| Microsoft November Patch Tuesday — carried | Up to 9.9 | Microsoft enterprise stack (Exchange, SharePoint, SQL Server, AD, LDAP, DNS) | One exploited (Outlook CVE-2026-124194 — see row above) | No change in-window. 1,094 newly released CVEs per Microsoft (Qualys counts 1,252 for the month; Tenable counts 1,093 — counting-method differences, flagged not resolved). Notable criticals, none reported exploited: SharePoint CVE-2026-120002 (9.9); SQL Server CVE-2026-116945 (9.8); Windows Search CVE-2026-116928 (9.8, zero-click); LDAP CVE-2026-119958 (9.8, zero-click, unauthenticated); Windows DNS CVE-2026-88218 (9.8); iSCSI CVE-2026-88234 (9.8); Exchange CVE-2026-118207 (8.8); AD DS CVE-2026-118146 (8.8); Schannel CVE-2026-119952 (8.1). Tenable notes the exploited Outlook CVE was discovered by a researcher using Claude. |
| CVE-2026-88771 — carried | 9.5 / 9.8* | Citrix NetScaler ADC and NetScaler Gateway | In the Wild (KEV) | No change in-window. Public Metasploit module (6.5.9, 6 Nov). Unauthenticated pre-auth command injection as root via attacker-controlled text written to appliance logs, later executed by maintenance script ns_monuploadd_err.pl — execution can lag the request by up to 24 hours. In KEV since 27 Sept. Fixed in 14.1-73.37 and 13.1-64.23. Patching does not remove existing implants. *CVSS v4.0 9.5 (Citrix) vs v3.1 9.8 (NVD). |
| CVE-2026-76443 — carried | 9.8 | Cisco Secure Email Gateway and Secure Email and Web Manager, 15.5 and earlier | In the Wild (KEV, sibling) | No change in-window. Public Metasploit module (6.5.9, 6 Nov). Rapid7 describes the module as an S/MIME/email-parsing SQL injection leading to unauthenticated root RCE — treat the CVE mapping as Rapid7’s, not Cisco’s. The actively exploited zero-day in the same family is sibling CVE-2026-76461 (9.8, in KEV since 14 Sept). |
| CVE-2026-97229 — carried | 10.0 | GitLab CE / EE | In the Wild (KEV) | No change in-window. Public Metasploit module (6.5.9, 6 Nov). CI/CD RegExp integer-overflow RCE as the git user; authenticated, requiring write access to a public project. In KEV since 29 Oct. |
| CVE-2026-119977 — carried | 8.1 / Low* | WordPress core 4.7.0–7.1.3 | In the Wild (attempts) | No change in-window. 240,000+ blocked attempts (peaking at 74,000+ on 30 Oct). Still no victim publicly named and no confirmed compromise. Fix 7.1.4 / backport; block web access to wp-content/upgrade/ before updating. Public PoC: lysophavin18/CVE-2026-119977 (GitHub). A second repo claiming to bypass the 7.1.4 fix holds no code, only a download link — flagged as a likely malware lure, do not download. *Wordfence 8.1 vs WordPress Low. |
| CVE-2026-107441 — carried | 8.8 (Zyxel) / 9.8* | Zyxel GS1900 series switches, ten models, web management interface | In the Wild (zero-day) | No change in-window. June fix added a check that ran after the unsafe copy in /cgi-bin/dispatcher.cgi, so June-patched firmware stayed exploitable. Payloads: Jitterly backdoor and Sixzut rootkit; suspected Chinese-speaking actor Kapibala/Red Heron (attribution via Binarly/Acronis, not independently verified). *Secondary analysis rates it 9.8. Not in KEV as of 7 Nov. Fix: 22 Oct firmware; change default credentials. |
| CVE-2026-109593 / 109594 — carried | 9.1 / 6.8 | SonicWall Global Management System 9.5.2 and earlier (end-of-life) | Not reported exploited in-window | No change in-window. 109593 critical (9.1): unauthenticated information disclosure via internal file reads. 109594 medium (6.8): high-privilege SQL injection. SonicWall will not patch — migrate to Network Security Manager. |
| CVE-2026-72898 — carried | 10.0 | Metabase 58+ (free and paid editions) | In the Wild | No change in-window. CSA Singapore warned 4 Nov that this SQL-injection flaw, fixed 6 Aug and on KEV since 11 Aug, is actively exploited and a PoC is public. Public PoC: 4minx/CVE-2026-72898 (GitHub), referenced in CSA alert AL-2026-158. Safe releases: 58.24, 59.21, 60.17, 61.11, 62.9, 63.5+. |
| CVE-2026-40217 — carried | 8.8 | LiteLLM proxy | In the Wild (KEV, added 18 Nov; LLMjacking) | CISA added it to KEV on 18 November (federal due date 2 December); the public proof of concept is a Nuclei template. User-ID rebind to PROXY_ADMIN. The flaw also affects Red Hat Ansible Automation Platform 2.6, OpenShift AI 3.3 and Lightspeed Core. Fix: LiteLLM 1.83.11+. Update LiteLLM, lock endpoints off the internet, rotate keys. |
| CVE-2026-98432 — carried | 9.8 | Linux kernel IPv4 TCP, 4.4+ | Research exploit only | No change in-window. Demonstrated exploit needs MPTCP; RHEL fixed kernels require reboot, Ubuntu was still in -proposed on 3 Nov. |
| CVE-2026-58115 — carried | 10.0 | Node-RED / Siemens IoT2050 | In the Wild | No change in-window. Port 1880 without adminAuth remains an active target. Siemens scope revision noted 8 Nov (more SIMATIC IoT2050 models). |
| CVE-2026-111497 — carried | 9.2 | Google Brotli; Cisco and Splunk products (Node2Shell) | In the Wild (KEV) | No change in-window. Scope as of 5 Nov: 19 Cisco products affected, dozens under investigation, about a dozen Splunk products also affected. Same Brotli root cause also affects ClickHouse. KEV federal due date fell 10 November, but per the KEV rule a deadline alone is not a new development. Qualys published detection/verification guidance for PHP2Shell and Node2Shell on 12 Nov (four non-intrusive QIDs, two intrusive verification QIDs) and reports 167 million PHP installations across 10,000+ customers; attackers probed PHP2Shell four days before a patch existed. |
| Citrix NetScaler November batch — updated | 9.5 (highest) | NetScaler ADC, NetScaler Gateway, NetScaler Console, Secure Access Client (Windows/Linux) | Not known exploited — no PoC reported | Status — the CVE records for the batch were published 13 Nov; no new development in this window. SecurityWeek counts 28 fixes, nine critical; SecurityOnline said four bulletins fixing 31 CVEs; The Hacker News said 12 flaws. The count is still unresolved between sources — do not pick one. Headline flaws: CVE-2026-107412 (9.5, memory flaw); CVE-2026-120868 (9.3, memory overflow in packet processing, affects all ADC/Gateway deployments); CVE-2026-120869 (9.4, improper authentication in NetScaler Console, unauthenticated bypass of authentication to access restricted resources — adjacent-network, needs Console IP access); CVE-2026-118268 (8.5, Secure Access Client for Windows privilege escalation to SYSTEM) and CVE-2026-118269 (Linux equivalent). Fixes: ADC/Gateway 14.1-73.57 and 13.1-64.39; Console 14.1-74.34. Citrix says it is not aware of any exploits for the announced vulnerabilities. Separately, CitrixBleed 2 (CVE-2025-5777) is still actively exploited — OffSeq honeypots recorded 316,950 exploitation attempts from 19 source IPs in the last 30 days, last seen 13 Nov 21:41 UTC (14 Nov 05:41 SGT), with DragonForce ransomware as the severe outcome in past cases. CVE-2026-107410 (9.3, heap overflow in NetScaler Push leading to denial of service; precondition is a PUSH load-balancing virtual server) was never named in any earlier edition. |
Showing all 54 watchlist rows.
Vendor, SaaS and service-provider exposure affecting financial organisations.
After its 48-hour deadline expired, n0n published customer marketing data for all three firms on 19 November. Each firm confirmed a third-party vendor incident: JPMorgan said third-party service providers were targeted and marketing data that may include customer names and email addresses was accessed; Visa confirmed an incident at an Asia-Pacific third-party provider with its own systems not involved; Mastercard confirmed a limited exposure from one benefit programme, contained, with no impact to its payment network. No card numbers, passwords or balances are in the published sets, and the Mastercard set looks like test data. No vendor has been named, and whether the three datasets came from one shared marketing platform is not established. The lesson for banks is the same one the Qubole campaign taught: ask every marketing and benefits vendor what customer data it holds, and treat a breach notice from one as your own exposure event. The Qubole link itself has no new development in-window: Wiz has still not verified n0n’s claim of responsibility, and Qubole / Idera has published no statement.
Forescout Research–Vedere Labs disclosed on 19 November that attackers abused a misconfigured Temenos Quantum Fabric deployment at a financial institution: an authentication service issuing privileged tokens was set to accept anonymous access and was internet-reachable, so attackers captured the anonymous token, exchanged it for an admin token without valid credentials, and planted FabricShell (a web shell running operating-system commands and database queries) and FabricShadow (a reconnaissance hook). Six more banks across Asia, the Middle East, Europe and the Americas share the same configuration and exposure — their compromise is not confirmed. Temenos attributed the behaviour to the anonymous-access configuration, released updates and guidance, and advised restricting internet access to admin and authentication services. The disclosure as read here rests on secondary reporting; the Forescout primary advisory URL was not retrieved verbatim in this run. Any bank running Quantum Fabric should check at once whether its authentication service accepts anonymous access and whether it is internet-reachable.
DHC disclosed about 19.11 million customer records leaked and Sharp disclosed up to 3.26 million COCORO STORE customers in notices published 18–19 November, pushing published counts across all KARTE client disclosures past 84 million. Plaid’s status page at 11:08 JST on 19 November still gave no KARTE restart date and said it was still sending affected companies their leaked data on request. MUFG Bank’s disclosure of about 6.5 million records was on 18 November. The KARTE incident is a supplier-breach story at scale: one platform’s data held the customer records of dozens of firms, and each firm is disclosing separately over days. Firms using customer-experience platforms should check what customer data those platforms hold and whether they have a restart and notification plan for a supplier breach.
Optimizely re-issued its incident PDF at the same URL as a corrected version dated 18 November (corrections to indicators in section 10 and clarification of the 16:06 Cloudflare block); its status page was unchanged as of 19 November and the incident is still listed unresolved. The scope estimate widened in a 19 November update: Zscaler reports more than 450 domains and more than 3,000 URLs serving the script, against the 178 organisations and 188 poisoned snippets previously counted, and about $7,040 has been traced to attacker wallets. Salesforce published its own OPTIMJACK response on 19 November, asking customers to check Marketing Cloud landing pages, Experience Cloud sites and Commerce Cloud storefronts. No statement from Kissmetrics or minute-ly has appeared. The fix is unchanged: hunt for 1688360[.]cc / track.js in Project JavaScript, block the domain, and remove the poisoned project content — an embedded fallback payload survives a host block alone.
Wiz’s 18 November update assesses that a client certificate and private key for Microsoft’s Engineering System Release Pipeline (ESRP) were compromised; a rejected publishing attempt from AWS was followed by success from Microsoft-owned IP space, and neither location identifies the actor. The malicious package versions 1.26.1 and 1.26.2 install info-stealing malware that exfiltrates environment variables and credential files and requests fresh identity tokens; the tooling identified itself as “Cairn”, an open-source AI penetration-testing framework. Wiz counted 50 captured exfiltration requests — four attacker tests, 29 sandboxes or scanners, and 17 requests representing three victim exfiltrations — not a verified victim total. Microsoft has still published no security advisory; its only statement reported so far is that it has “made some repositories private while we investigate malicious content”. Both malicious versions were removed within about an hour; 1.26.0 is the last clean version. Any firm that installed 1.26.1 or 1.26.2 should remove the package, rotate Azure, Git, Docker and Hugging Face credentials, and review GitHub Actions and Azure DevOps token issuance.
A 19 November piece recaps Fireworks’ 17 November statement that it identified two separate incidents, both contained and remediated; Uber notified about 150,000 drivers. Nothing new beyond the 17 November statement has appeared in-window. The standing advice holds: firms using third-party AI platforms should treat a breach notice from an AI vendor as their own exposure event — check which of their API keys, tokens and data the platform held, and rotate them.
Hesse data-protection commissioner Alexander Rossnagel opened an investigation into whether Deutsche Bank implemented appropriate technical and organisational measures, following the leak of about 15 million internal records disclosed this week. CEO Christian Sewing said on 19 November that clients were calm and the bank sees no business impact; shares fell more than 6% initially. The bank says it proactively informed the authority in October. The leak itself is a disclosure of internal records, not a confirmed system breach, and the files’ contents are reported through the journalists’ investigation, not independently verified here.
The latest reporting remains 16–17 November (Expedia’s confirmation on 12 November; n0n’s claims on 13 November). Nothing new on 19 November. The standing position holds: Wiz documented the campaign on 9 November (updated 11 November) and has not verified n0n’s claim of responsibility; Expedia has not named Qubole; Qubole / Idera has published no statement.
No new vendor or researcher update on 19 November; the latest is the second SCKit wave on 13 November. The fix remains the same: upgrade ScreenConnect technician clients to 26.6.5 or later (CVE-2026-84869, 9.9, exploitation confirmed) or remove the TransferFiles permission — the server patch alone does not fix it — and upgrade N-able N-central to Hotfix 3. Firms should ask their providers whether they run ScreenConnect with the RMM+ extension and whether technician clients are on 26.6.5 or later, whether they run N-central on Hotfix 3, and hunt for the RelayTrap task, the TrapWatch task, the update_<8 hex>.bat pushes and the N-central IOCs in the Indicators section.
The Financial Supervisory Service convened all domestic virtual-asset business CISOs (exchanges and custodians) on 19 November to review security posture and incident response, following the Gopax / Bithumb / ProBit incidents. No outcome, statement or FSS press release from the meeting itself surfaced in Korean-language searches. No new development beyond the 15 November disclosure (70-hour-40-minute disclosure delay, 11 data categories). Passwords, resident registration numbers, bank account numbers, wallet addresses and transaction IDs given during support contacts may have leaked, while Coinone says its own account, trading and wallet systems were not directly breached.
Osaka City’s disclosure of 1,078,713 email addresses leaked via the Arara email-delivery platform was covered in the 15 November edition; no new development appeared in this window. The platform’s 6 November third report put the total at 81,329,467 email addresses (deduplicated) across up to about 380 client companies; another source cites about 77 million — the two totals conflict, do not merge them. The same platform was abused 28–30 October to send up to about 7.13 million phishing emails impersonating a credit-card company.
The open database that exposed 125 million+ SMS messages, including Swedbank account details and one-time codes, has no new development in this window. Swedbank has stopped using the provider and opened an investigation; CERT.LV is investigating. No evidence anyone other than researchers read the data; no confirmed account takeovers or financial loss.
Fraud, infostealer and social-engineering developments affecting financial firms and their customers.
Gurugram Cyber Police arrested two men in Ludhiana on 15 November over an alleged attack on PayU Payments’ cloud servers, following a complaint filed by PayU on 9 November. Attackers gained access to three cloud servers, manipulated the company’s payout API and carried out 265 unauthorised transactions involving 156 bank accounts, transferring ₹25.67 crore (about US$2.9 million) without authorisation. PayU’s security team detected unusual API activity around 23:30 on 5 November, after the transfers had taken place; the company’s cloud service provider had also flagged suspicious access. Police have frozen around ₹1.5 crore of the stolen funds. How the attackers obtained access to the cloud servers, whether security controls failed and whether any merchant funds were affected have not yet been established; an audit report is awaited. The operational point is the same as the NBFC loan-diversion pattern: a disbursement API is a standing instruction to move money, and an attacker who reaches it does not need to break customer authentication.
A secondary outlet updated on 19 November puts the wider loss at about ₹105.67 crore across at least 14 NBFCs, with about ₹19.12 crore traced or lien-marked, and says investigators detected indications AI tools were used to automate parts of the attacks. That figure conflicts in scope with the better-sourced figures from Delhi Police via the Indian Express — 12 Delhi-NCR NBFCs, more than ₹65 crore over three months, 20 arrests, ₹18 crore seized, more than 150 FIRs nationwide and more than ₹300 crore in total — so the two are unreconciled, not a confirmed revision. No new arrests, charges or loss figures were found in-window. Separately, Pimpri-Chinchwad Cyber Police issued an advisory dated 18 November (reported 19 November) warning banks, NBFCs and fintechs of AI-assisted fraud misusing fund-disbursement systems and APIs for unauthorised transactions “worth crores”, urging API inventory hygiene, disabling unused APIs, prompt CERT-In reporting and RBI DAKSH logging, and coordination with the Indian Cyber Crime Coordination Centre for bulk freezing. No institution was named and no loss total was given.
Hunt.io research, reported by CyberPress on 19 November, found an exposed attacker directory on a server in the Netherlands holding 1,692 files totalling about 1.11 GB of stolen data, malware tools and operational logs. The logs recorded 603 distinct implant IDs across 469 hostnames between 12 October and 8 November 2026. At least 57 victims had a US bank or government login in the stolen records; Chase appeared in 31 victim datasets, Capital One in 20, Citi in 17 and Wells Fargo in 15 — figures showing where customers had accounts, not where attackers successfully signed in. One decrypted LastPass vault held 2,880 logins, including Chase, Wells Fargo, the IRS, Medicare and PayPal, plus 17 passkeys; the attackers also cracked two 1Password vaults and copied a KeePass database from a developer workstation. In two Microsoft 365 cases the attackers reused stolen browser sessions to obtain Microsoft Graph access tokens without a fresh multi-factor authentication prompt. The researchers found no evidence that the banks or federal agencies themselves were breached — the credentials were stolen from victims’ own infected PCs. Whether any of the stolen bank credentials have been used for fraud or account takeover is unconfirmed. The control lesson is on the endpoint: infostealer logs are the raw material for account takeover, and session-token theft defeats multi-factor authentication without touching the bank.
An article dated 20 November (00:29 CET / 07:29 SGT) reports an Android campaign under the “CyberDakait” label targeting Indian banking customers: reportedly at least 17,960 devices compromised, 7,611,443 SMS messages intercepted, seven malware families grouped under the label, with financial-data theft plus crypto-mining reported. The article itself stresses the theft path, victim impact and losses are unconfirmed, no victim bank is named and no single root cause is established. Treat as an unconfirmed lead only: no bank has confirmed an incident, and no loss figure is verified.
Singapore Police said on 19 November that 235 people aged 16 to 93 are being investigated after a two-week islandwide operation (5–18 November) covering more than 536 scam cases and about $4.2 million in losses — e-commerce, phishing, job, government-official impersonation, investment and lucky-draw scams, with money-mule offences included. No bank system compromise is alleged. It is an enforcement result, not a new fraud technique, but the volume shows the impersonation pipeline that feeds on leaked customer data is still running at scale.
The n0n publication gives phishers real names, emails, phone numbers and travel and dining booking details for premium Visa and JPMorgan Reserve customers — expect messages citing cardmember programme and benefits-plan details, and warn premium clients to verify any message through known channels. Rakuten Securities’ leaked login IDs, withdrawal bank accounts, income and asset data remain impersonation material, and a fake “compensation policy” screenshot is circulating — Rakuten has announced no compensation. The KARTE disclosures (now more than 84 million records) are impersonation material for customers of MUFG Bank, DHC, Sharp and the other affected firms. The KG Eduone leak gives attackers real savings-bank staff names, job titles, phone numbers and email addresses — expect calls impersonating bank staff, and remind customers that banks never ask for account passwords, security-card numbers, one-time passwords or authentication codes by phone or text. PayU’s payout-API theft and Bell Financial’s incident will be cited in impersonation messages to merchants and clients. The Wellington IT payee data (IBANs, account numbers, sort codes) enables precisely targeted phishing against credit-union members, using real payee names and payment references. The Comerica and HSBC Taiwan forum claims are unverified and should not be treated as breaches, but the claims themselves will be cited as phishing bait. Coinone’s leaked passwords, resident registration numbers, bank account numbers, wallet addresses and transaction IDs remain voice-phishing and account-takeover material. The FSI special audit began on 16 November — expect phishing citing the FSI or the evaluation process, using the real assessment-staff contact details and firm names from the leak. The SBI FIR names a group, “Siriusim”, and confirms loan-application and account-opening data was in the stolen set. Monex’s 1.13 million email addresses and AXA’s about 597,000 accounts remain impersonation material, as do KOMSCO’s chak-user details (KOMSCO warned of voice-phishing and smishing; a self-check function opened in the chak app from 17 November) and KBIZ’s Yellow Umbrella details (KBIZ warned of impersonation and loan fraud). Arara’s leaked shareholder emails are live phishing bait, and the same platform was already abused to send about 7.13 million phishing emails impersonating a credit-card company. The Tesouro Nacional leak (names, CPFs, Pix keys) and Caixa’s leaked CPFs are impersonation material for 7.6 million-plus people. SMBC Nikko’s leaked holdings and account details are impersonation material for shareholders. Mirae Asset’s leak gives attackers 3,800 customer account numbers plus names, emails and mobiles of 35,255 employees. Binance, Mercado Bitcoin, Gopax and Nubank/Mercado Pago outage messages remain live phishing bait, and the Microsoft 365 outage credential-phishing campaign (sender domains moments-notice.com, planetmaison.com, newbooker.com; click-tracking reallifeaccounting[.]com) is confirmed by Microsoft — verify via official status pages only. Autopay customers in Poland should watch for phishing citing the payment processor, after customer notifications named suspected access to email, phone and payment-card data. Watch for Rs 1 test transactions on compromised accounts.
No new DDoS-extortion item targeting a financial institution was found in this window. The one new mobile-malware item is the thinly sourced “CyberDakait” report above, which is unconfirmed. The CSC Telecom open database (125 million+ SMS messages including Swedbank one-time codes) was published on 11 November, outside the window, and has no new development.
Customer allegations of thefts from Bank of Khartoum accounts via the Bankak mobile banking app remain reported / unconfirmed: more than 50 traders and customers in West Kordofan claim millions of Sudanese pounds were stolen, with individual reported losses of SDG 18–45 million. The bank has still not responded to comment requests, and no technical evidence that the app itself was breached has been published. Treat as an allegation, not a confirmed bank breach.
Techniques actually published for this edition’s actors and key vulnerabilities. No MITRE mapping is shown where the source published none.
| MITRE Technique ID | Technique / behaviour | Seen in this edition | BFSI relevance / detection note | Status |
|---|---|---|---|---|
| — none published | Unauthenticated authentication bypass in VMware vCenter Directory Service; SSO administrator account creation, directory-database exfiltration and SAML signing-key theft (CVE-2026-123890, ARMORY CASTLE) | Google Threat Intelligence Group / Mandiant, 19 Nov | No MITRE IDs were published. Detection rests on patching vCenter (9.1.1.0200 / 9.0.2.0200 / 8.0 U3m) and hunting for unauthorised SSO administrator accounts created since early September and for forged SAML assertions. There is no workaround. | No mapping published |
| — none published | Anonymous-access token exchanged for an admin token on a bank’s digital-banking middleware, then web-shell and reconnaissance-hook upload (BankWeaver / Temenos Quantum Fabric) | Forescout Research–Vedere Labs, 19 Nov (via secondary reporting) | No MITRE IDs were published. Detection rests on checking whether the Quantum Fabric authentication service accepts anonymous access and whether it is internet-reachable, and on hunting for FabricShell / FabricShadow components. Six more banks share the exposure; their compromise is not confirmed. | No mapping published |
| — none published | Payout / disbursement API manipulation to move money without breaking customer authentication (PayU; NBFC loan-diversion pattern) | MediaNama / Gurugram Cyber Police, 19 Nov; Delhi Police via Indian Express, 18 Nov | No MITRE IDs were published. Detection rests on monitoring disbursement flows for destination-account changes after approval, bursts of unauthorised transactions against approved loans, and unusual payout-API activity outside business hours. PayU detected the activity only after the transfers had taken place. | No mapping published |
| — none published | Session-handling remote code execution in PHP (session serializer mismatch, TOCTOU session files, PHP-FPM request smuggling) and unauthenticated authentication bypass in Splunk via an empty clustering shared secret (CVE-2026-76536) | PHP project security release, 19 Nov; Splunk advisories, 19 Nov | No MITRE IDs were published. Detection rests on applying the PHP security releases (8.2.36 / 8.3.37 / 8.4.28 / 8.5.13 / 8.6.0) and the Splunk update (10.6.1 / 10.4.5 / 10.2.9 / 10.0.12 / 9.4.17), or setting a strong unique pass4SymmKey for Splunk. No exploitation of either has been reported. | No mapping published |
| — none published | Infostealer credential and session-token theft from infected PCs; stolen browser sessions reused to obtain Microsoft Graph access tokens without a fresh multi-factor prompt (Hunt.io) | Hunt.io research via CyberPress, 19 Nov | No MITRE IDs were published. Detection rests on endpoint infostealer hunting and on watching for anomalous Microsoft Graph token use from customer and staff machines. The bank logins found (Chase, Capital One, Citi, Wells Fargo) were stolen from victims’ own PCs — no bank was breached. | No mapping published |
| — none published | AI-generated cyber campaign against the European Commission (misuse of a publicly available AI tool); reported ARTEX use unconfirmed | European Commission statement, 19 Nov; Politico / The Record, 19 Nov | No MITRE IDs were published, and the Commission named no tool, entry point or affected application. The ARTEX link rests on press reporting of CERT-EU’s identification and is unconfirmed by the Commission — do not present it as established. Detection rests on the standing AI-agent and tool-use monitoring in the AI section. | No mapping published |
| — none published | Publication of stolen marketing data after a ransom deadline expires (n0n — Visa, JPMorgan Chase, Mastercard) | Reuters / Cybernews, 19 Nov | No MITRE IDs were published. Detection rests on warning premium clients about impersonation using the leaked names, emails, phone numbers and booking details, and on watching for staff-targeted phishing using the exposed employee addresses. No payment data is in the published sets. | No mapping published |
| T1078.004 | Valid Accounts: Cloud Accounts — stolen vendor IAM credentials used for cross-account AssumeRole to steal secrets and S3 data (Qubole campaign) | Wiz research, 9–11 Nov, via The Hacker News; ransomware.live notices, 18 Nov | The technique is already recorded as confirmed on the n0n profile page for the Qubole campaign; it is not a new sighting in this window, and the link between the Qubole campaign and n0n is unverified. Detection rests on hunting CloudTrail for AssumeRole calls from non-AWS addresses and for the quiam-* role session names — see Indicators. | Context only |
Recurring-TTP watch: the technique that keeps recurring in this edition is trusted third-party access — marketing vendors holding premium-cardmember data for three firms at once (n0n), a data-platform provider’s credentials opening many customers (Qubole), a bank’s own digital-banking middleware reached through a misconfigured authentication service (BankWeaver), a payment provider’s cloud servers and payout API (PayU), and tax documents held by an adviser (KPMG Canada). Alongside it, the control plane is the target: vCenter’s directory service and SAML signing key, and Splunk, the SIEM itself. No technique was published with a MITRE ID for any of this edition’s actors or vulnerabilities except the Qubole item, which is context only — that gap is stated, not filled.
AI as a target, AI as a weapon or accelerator, and AI as a risk and control problem for BFSI.
Microsoft agent-framework PyPI hijack — Wiz ties it to a compromised release-pipeline credential (update, 19 Nov). Wiz assesses that a client certificate and private key for Microsoft’s Engineering System Release Pipeline were compromised to publish the malicious agent-framework versions 1.26.1 and 1.26.2. The stealer harvested environment variables and credential files, including LLM API keys and Hugging Face tokens, from AI-agent development environments, and requested fresh identity tokens from GitHub Actions, Azure DevOps and Azure. The tooling identified itself as “Cairn”, an open-source AI penetration-testing framework. Only three victim exfiltrations were observed in the sample Wiz could read — not a verified victim total. Microsoft has still published no security advisory.
LiteLLM KEV attribution is contested (in-window coverage, 19 Nov). Google Threat Intelligence Group attributes an LLMjacking intrusion — attackers who bought ordinary LiteLLM virtual keys escalating them to proxy-admin credentials — to CVE-2026-40217 in its 2 November report and 16 November AI Threat Tracker; Corgea disputes the attribution, saying the key-escalation technique is a separate bug fixed in 1.84.0. The attribution is contested; do not present it as settled. The flaw itself (CVSS 4.0 5.9) was added to CISA KEV on 18 November and is covered in the watchlist.
No new exposed-LLM-endpoint or LLM/API-key theft incident was found in this window.
European Commission confirms an AI-generated cyber campaign (confirmed, 19 Nov). The Commission confirmed it responded to a detected AI-generated campaign — misuse of a publicly available AI tool — detected on 22 October and confirmed after initial investigation on 26 October. Several applications were taken down temporarily as a precaution; most have been restored. No attribution, no ransom demand, and no tool, entry point or affected application was named. A report that CERT-EU identified ARTEX use is unconfirmed by the Commission. AP called it the first time the EU has publicly reported a cyberattack using an AI tool. This is an observed AI-tool attack disclosure, not a forecast.
No new AI-scaled phishing, vishing or deepfake campaign against a named bank was found in this window. The Coinbase deepfake-Elon-Musk livestream item (AI defences stopped about $4.7 million in transfers) was covered in the 19 November edition; no new development appeared in-window. A 19 November analysis piece calls autonomous offensive frameworks “the third wave of AI-driven attacks” and recaps ARTEX, PentAGI and Cairn intrusions — that is analysis and opinion, not a new incident, and its forecast of stealthier self-hosted operations within three to six months is a forecast, not observed activity.
Singapore MAS warns against giving AI agents bank login credentials (confirmed, 19 Nov). MAS cautioned against giving AI agents login credentials to personal bank accounts and said Singapore is developing a common protocol so banks can verify AI agents before allowing them to transact on customers’ behalf, defining what agents are authorised to do and their limits. Context: the CSA / GovTech / IMDA joint advisory AD-2026-015 (6 November) on personal AI agents and one-time passwords remains the standing guidance.
UK regulators consult on cyber resilience, with AI and quantum effective practices (confirmed, 19 Nov). The Bank of England, the Financial Conduct Authority and the Prudential Regulation Authority published a joint consultation on cyber risk and resilience management for financial firms, open until 26 February 2027, aligned with the NCSC’s Cyber Assessment Framework, plus separate effective practices on AI and quantum risks. Context: October 2026 saw the second-highest monthly number of cyber incidents reported to the FCA in five years, 52% above the previous 18-month peak in May 2025, and the Bank’s Systemic Risk Survey cites cyber as a top systemic risk.
Korea’s FSS asks firms for AI-industry exposure (reported, 19 Nov). South Korea’s Financial Supervisory Service asked major banks, securities firms and insurers to disclose direct and indirect AI-industry exposure (investments and loans); described as preventive, with no aggregate figure or deadline published. It is a regulatory signal, not an incident.
Singapore CIDeX exercise features AI-agent attack simulation (confirmed, 19 Nov). The Critical Infrastructure Defence Exercise (CIDeX), 17–20 November, features a coordinated “swarm” of AI agents simulating large-scale AI-enabled attacks against IT/OT testbeds, with more than 250 participants from 44 organisations, covering all 11 critical information infrastructure sectors including Banking & Finance. It is an exercise, not an incident — but it shows regulators now test defences against AI-agent attack swarms as a matter of routine.
Visa: AI is accelerating traditional fraud, not creating new categories (report released 17 Nov; covered in-window 19 Nov). Visa’s Fall 2026 Biannual Threats Report finds incidents involving third-party agents rose 62% against the prior six months, with third parties linked to about 56% of compromised payment credentials identified. The control problem is the same one this edition keeps returning to: the agent, the supplier and the platform are now part of the attack surface, and the fraud they enable looks like ordinary fraud, only faster.
Notable indicators from this edition, and prioritised steps for SOC and vulnerability-management teams.