Broadcom patches vCenter zero-day exploited since September | Magica
Broadcom patches vCenter flaw exploited since September, Mandiant says
Editorial Team
••📖6 min read
A critical authentication bypass gave a suspected China-linked espionage group access to VMware's virtualization control plane. Mandiant says the campaign stole signing keys and virtual-machine data, making compromise checks essential alongside patching.
Broadcom has patched CVE-2026-123890, a critical authentication bypass in VMware's vCenter infrastructure-management software.
Mandiant says Armory Castle exploited it in early September, after entering through a separate BeyondTrust vulnerability.
The documented attack reached signing keys, ESXi hosts and sensitive files inside virtual machines.
Broadcom released patches on November 19 for a VMware vCenter Server flaw that Google's incident-response and threat-intelligence business, Mandiant, says a suspected China-linked espionage group exploited in early September 2026. The documented attack reached the software controlling virtualized infrastructure, then extended into credentials and files inside the virtual machines it managed.
Mandiant’s reconstruction of Armory Castle’s lateral movement and vCenter compromise. Source: Google Threat Intelligence Group and Mandiant. Source: Google Cloud.
A management server becomes a route to guest data
vCenter configures, controls and monitors ESXi hosts, the hypervisors that run virtual machines in VMware's vSphere platform. Broadcom, the semiconductor and infrastructure-software supplier that acquired VMware in November 2023, supplies the fixes. Its security advisory rates the flaw at 9.8 on the CVSSv3 severity scale and says an attacker with network access to vCenter can exploit the VMware Directory Service to bypass authentication. It lists no workaround.
Mandiant joined Google Cloud after Google's September 2022 acquisition and reported this vulnerability to Broadcom. Its researchers and Google's threat-research team, Google Threat Intelligence Group, describe forensic evidence in their November 19 analysis. Broadcom's advisory itself does not state that exploitation occurred.
The management company supporting Japan's Shonan Beauty Clinic network disclosed stolen customer passwords stored without encryption and a separate booking-service incident. Treatment-related submissions were also exposed, while the affected population and financial impact remain under investigation.
Editorial Team
Google attributes the activity to Armory Castle, formerly tracked as UNC5221, a suspected China-linked espionage actor it has tracked exploiting edge and security appliances for initial access since 2023. The newly documented vCenter exploit extends that record into using a zero-day after entering the network. The attribution remains Google's assessment, rather than an identification established in Broadcom's notice.
In the documented campaign, the group first exploited a separate zero-day in BeyondTrust Remote Support, CVE-2026-102764. It moved from that device to an internal Linux appliance lacking endpoint detection and response tooling, escalated privileges and targeted vCenter from inside the network. Internet exposure of vCenter was therefore not a prerequisite for this attack path.
Mandiant says the group's BONESAW utility exploited the new flaw to establish an administrator session, create a single sign-on administrator account and copy vCenter's directory database. The vulnerability exposes a shared authentication session key through the directory service's remote-procedure-call interface on TCP port 2012.
The stolen database supplied material for a further attack on identity controls. BONESAW can extract SAML signing keys, used to sign authentication tokens that vCenter trusts. Another utility, SAWSNAKE, uses that material to forge tokens, obtain administrator sessions and bypass multifactor authentication, according to the analysis. The researchers describe SAWSNAKE as a Go adaptation of an existing Python vCenter SAML login tool: the campaign combined a newly exploited flaw with an established token-forgery technique.
Mandiant subsequently observed the attackers stealing vCenter cryptographic material and database credentials, accessing ESXi hosts, adding an SSH key for persistence and extracting files from virtual-machine disks. Those files included Windows Active Directory database material and registry hives. The researchers traced staged archives back to the Remote Support appliance but described onward exfiltration from there as likely.
Mandiant observed September exploitation against a server already patched for the separate July vulnerability. Dates mark reported events; September timing is approximate. Source: Google Cloud.
The July patch did not fix this vulnerability
Mandiant says BONESAW ran successfully against a vCenter server already patched for CVE-2026-59309, a different authentication bypass disclosed on July 29. Although both vulnerabilities involve obtaining a shared session key without knowing a password, the earlier issue used the LDAP interface; the newly disclosed flaw uses the directory service's DCE/RPC interface and predictable timestamp generation.
That distinction matters for administrators who addressed the earlier advisory: the documented target's July fix did not prevent the September compromise.
The recovered BONESAW binary carries a July 13 build timestamp, suggesting development before the earlier flaw's disclosure. Mandiant cautions that Go build metadata can be modified. The timestamp is consequently weaker evidence of when the tool was developed than the execution logs are of September exploitation.
The researchers interpret the campaign's two zero-days and platform-specific malware as evidence of substantial resources and technical expertise. Mandiant chief technology officer Charles Carmakal said the company investigated multiple incidents involving exploitation of this flaw, in a LinkedIn statement cited in a report on the disclosure. Neither that account nor Google's technical analysis supplies a victim count or a measured share of affected deployments. Multiple investigations establish activity beyond a single incident without showing how widespread compromise is.
Broadcom’s response matrix for CVE-2026-123890 lists branch-specific fixes, the vCenter 7.0 support route and guidance for Telco Cloud deployments. Source: Broadcom.
Patch the affected branch and investigate prior access
Broadcom's response matrix lists these remediation paths for the vCenter flaw:
Deployment branch
Listed fix or support route
vCenter 9.1.x.x in Cloud Foundation or vSphere Foundation
9.1.1.0200
vCenter 9.0.x.x in Cloud Foundation or vSphere Foundation
9.0.2.0200
vCenter 8.0
8.0 U3m
vCenter 7.0
Contact Broadcom Support with an extended support contract
VMware Cloud Foundation 5.x
5.2.5
The advisory also covers affected Telco Cloud deployments through product-specific guidance. Its other four vulnerabilities involve SQL injection, authenticated command injection and local privilege escalation in SDDC Manager, plus a token-signature bypass in VMware Tools for Windows. Those require authenticated or local privileged access; they should not be confused with the unauthenticated vCenter flaw.
The support route is especially consequential for older deployments. BONESAW compromises vCenter 7.x and 8.x appliances, according to Mandiant. Version 7.0 reached the end of general support in October 2025 and has no public patch for this flaw, a report on the disclosure notes.
Broadcom treats the vCenter fix as an emergency update and says network restrictions can reduce exposure, according to a report on its guidance. The same report says the flaw affects vCenter whether or not Enhanced Linked Mode is enabled. Network restrictions do not replace the patch, and the observed attack from an internal appliance makes internal access controls relevant too.
The remaining question for affected organizations is whether an attacker arrived before remediation. Mandiant's analysis and hardening guidance recommends immediate patching and publishes indicators and detection rules to support investigation. It also warns that the default management-interface firewall lacks granular port controls and remote logging of denied connections and shell activity. It recommends stricter network rules and centralized audit collection, with rigorous testing and formal change control before deploying its hardening tool in production.
The observed creation of administrator accounts, theft of signing keys and installation of persistence give investigators concrete reasons to examine prior access alongside the update. Installing the fixed version addresses the vulnerability; determining whether the environment's identities and workloads were already compromised requires evidence from the environment itself.
N-able says an intruder reached a server supporting Mail Assure and SpamExperts Hosted Cloud. Mail delivery and filtering continued, but customers lost management access as the company revoked tokens and worked to restore services securely.
Microsoft is testing an optional Edge mode that blocks WebGL and WebGPU to reduce opportunities for browser exploitation. An early hands-on test found that restoring graphics for a trusted site also removed its broader Enhanced Security protections.
PHP’s November 19 security releases address seven critical flaws, including unauthenticated code execution through reused backend connections. Patching takes priority, but applications that retain session upload progress must also check shared storage and serialization settings.
TfL has shut its Oyster photocard website after a breach that may have exposed around 265,000 customers’ account details, including security answers. Existing cards still work, while the attack method and restoration timetable remain undisclosed.
Around 40 Irish credit unions are implicated in a saved-payee data breach at software supplier Wellington IT. Missing logs leave the victim count unresolved, while a minister is seeking answers about the disclosure delay.
JPMorganChase says it believes attackers accessed marketing data at outside service providers, potentially exposing customer names and email addresses. The bank says services are unaffected, but has not identified the vendors or disclosed the attack date.
Apache confirms exploitation of a Roller code-execution flaw that survives disabling XML-RPC in its administration panel. Attacks on the retired Axis 1.x library also expose enterprise applications, making product-specific mitigation and compromise checks urgent.
StepSecurity’s Elephant-led Series A will fund broader protection for developer machines, package downloads and build pipelines, alongside planned defenses for AI-agent sandboxes. Its challenge is turning open-source adoption into a larger commercial platform while proving the extra controls earn their place.
Nvidia says it has found no indication its own systems were affected after The Gentlemen ransomware group listed it on a leak site. The investigation concerns a possible incident at a partner, while the gang has supplied no public proof.
Senators from both parties plan to delay the CISA nomination of Palantir security chief Michael LoSapio after his Jan. 6 denial, adding a confirmation obstacle as the agency faces staffing losses and funding decisions.
Sean Cairncross is leaving as national cyber director, with Pentagon CIO Kirsten A. Davies reportedly selected to succeed him. The transition puts Senate confirmation at the center of a cybersecurity leadership reshuffle.
Windows baseline protection combines existing PC safeguards into a verifiable pass-or-fail result. Microsoft defines the standard; developers choose whether failure means a warning, fewer features or blocked access.
F5 lists fixed releases for a critical nginx HTTP/2 flaw, while the wider NGINX2Root disclosure covers five independent nginx attack paths and a separate Linux kernel bug. Remediation depends on configuration, supplier patch levels and replacing vulnerable running processes.
Tyler Robert Buchanan received six years in U.S. prison for a phishing and SIM-swapping conspiracy that stole at least $11.6 million from individuals. He is the last of five defendants sentenced; restitution remains pending.
The new Threat Disruption Center makes TRM Labs its main technology partner following a reported $94.66 million sole-source award for Homeland Security Task Force investigations. A rival lost its court challenge, but the launch leaves the center’s share of that spending and its performance benchmarks undisclosed.
Fandango says payment details from certain October 8–11 transactions may have been exposed. State disclosures provide only a partial affected count, while the company has not explained how the checkout page was compromised.
Norway’s state-controlled oil and gas group is working to neutralise an attack on some IT systems. Employee services may be disrupted, but Equinor says operations and production remain unaffected and it has found no indications of data theft.
Tor urges relay operators to install 0.4.9.17 as a continuing DDoS attack strains its privacy network. A post-upgrade operator account describes further disruption, while Bitcoin connectivity measurements show partial recovery with important limits.
Malicious code in Optimizely website-testing projects could swap cryptocurrency addresses and skim payment cards. Optimizely says it has fixed the vulnerability and removed the code, but visitor exposure and losses remain unquantified.