IPCRIME | November 21, 2026. Researchers at Palo Alto Networks Unit 42 have published new evidence of campaigns by loosely connected cybercriminal groups known in security research as ‘Com’-affiliated actors. The report differentiates several clusters that can share infrastructure and tactics while conducting separate operations. Two newly detailed clusters, Vortex (CL-CRI-1264) and Cinder (CL-CRI-1274), were observed from September 2026 and used voice phishing and stolen sessions to access cloud data and demand payment.
The research is particularly relevant to IPCRIME’s Underground Files: a name used by an extortion actor, an incident-response firm’s tracking label and the identity of a person behind a breach are three different claims. Evidence of similar tactics is not enough to merge every group into a single criminal organization, nor does it establish any particular individual’s guilt.
Who Unit 42 is tracking
Unit 42’s November report distinguishes Bling Libra, also called ShinyHunters, and activity clusters it labels Redact, Pink, Falcon, Helix, Vortex and Cinder. Those names represent distinct analyst classifications, not a formal membership roster of a unified group. Some reportedly share approaches, social contacts and hosting patterns. Their boundaries can change as investigators learn more.
The term ‘Com’ in the report describes a broader ecosystem of affiliated or overlapping cybercrime activity. A large ecosystem should not be treated as a single actor with control over every intrusion that uses a fake IT-support page or an adversary-in-the-middle phishing site.
Vortex: identity deception directed at healthcare
Unit 42 says its Vortex cluster primarily targeted healthcare, pharmaceutical and medical organizations from September. Reported access began with callers impersonating internal technology support, directing staff to company-branded pages that proxied legitimate sign-ins. Researchers observed accounts in which an unauthorized MFA device was enrolled minutes after initial authentication, preserving account access even when the original login looked routine.
The same reporting describes scripted attempts to collect Microsoft 365 mailbox, OneDrive and SharePoint content, followed by internal messages asserting data theft and sending recipients toward negotiation sites. Those observations support a cloud data-theft and extortion pattern, but the public report does not establish a definitive victim count or connect all healthcare breaches to Vortex.
Cinder: persistent sessions and internal messaging
Unit 42 describes Cinder as a separate September cluster that reused persistent authentication sessions to avoid fresh multifactor prompts, then used scripted access to cloud collaboration content. Researchers also observed extortion messages sent through compromised accounts and internal platforms. A session remaining active after an account’s password changes is an important risk, but it does not mean every user who selects ‘keep me signed in’ is compromised.
The disguised browser-extension problem
Another reported cluster, Helix, was associated with a malicious Chromium extension masquerading as ‘Google Chrome’. Unit 42 says the tool, installed outside the official browser extension store, could read cookies and browser storage, capture screenshots, observe input and abuse browser debugging capabilities. That is evidence from analyzed samples. Unit 42 characterizes the attribution of the extension’s development to that activity cluster as a moderate-confidence assessment, not a definitive identification of the programmer.
What the geographic figures do and do not show
The researchers’ tracked phishing subdomains overwhelmingly targeted U.S. organizations in August and September. The Australian share of observed targeting increased to approximately 13% in October, with healthcare overtaking finance during the same general period. These statistics reflect Unit 42’s visibility into tracked infrastructure; they are not a census of all cybercrime in the world or a count of successful breaches.
Defensive lessons from the evidence
- Use independent callbacks to verify unexpected help-desk calls before taking account-recovery or MFA-enrollment steps.
- Require phishing-resistant authenticators for sensitive cloud accounts and monitor newly enrolled authenticators and recovery methods.
- Review session revocations, application grants and suspicious Microsoft 365 data exports during incident response; a password change alone may be insufficient.
- Restrict browser extensions through approved software policies and inspect unexpected installations that request broad browser privileges.
- Preserve evidence without publicly distributing stolen documents, account tokens or victims’ identifying information.
Sources and editorial independence
Palo Alto Networks Unit 42 — Dialing for Data: Inside the Com’s High-Profile Heists (published November 20); IPCRIME independent Threat Actor Profiles.
This article paraphrases public threat research, separating observed activity from attribution hypotheses. IPCRIME has no law-enforcement partnership and has not independently accessed the attackers’ infrastructure.