Cyber Roundup: Week of November 9

Here are the main cyber stories you missed last week.

Key takeaways

  • Maileva: A La Poste group platform that businesses and agencies use to send mail exposed the names and postal addresses of 22 million people who never chose the service.
  • N-able: A researcher published a new N-central exploit chain after N-able fixed his earlier flaws. Attackers were compromising servers about five hours after the proof of concept appeared, and Huntback reports that 19 victims had applied the latest patches.
  • Check Point: Attackers chained PHP2Shell with CVE-2026-123039 to reach root on gateways, then ran a wiper on a VMware ESXi host. Patch, then hunt for the webshells Check Point Research lists.
  • Microsoft: Patch Tuesday fixed roughly 1,090 CVEs. The one exploited zero-day is an Outlook bug scored 6.5 that can hand attackers a victim’s NTLMv2 hash.
  • Operation Endgame: Europol and partners seized 235 servers and 257 domains tied to Gootloader, CastleRAT, and StealC, and identified credentials for 6.4 million victims. No arrests were reported.

1. Maileva: A mail platform that 22 million people never chose exposed their names and postal addresses

The headline: Maileva, the digital correspondence platform run by Docaposte inside the La Poste group, says an attack dating to mid-September 2026 exposed the contact details of 22 million people, as Il Sole 24 Ore reported on November 15. The exposed data is first name, surname, and postal address. La Poste Group confirmed that affected client companies have been notified, that the incident was reported to the CNIL, and that a complaint was filed. The Paris prosecutor’s cybercrime division opened an investigation into an attack on an automated data-processing system and gave it to the specialist unit BL2C. No attacker has been named and no ransom demand has been reported. Maileva handles mail for private companies and government agencies, so the 22 million are the contacts of its clients, not Maileva’s own users.

What we’re actually watching: Most of the 22 million never signed up to anything. Their organizations outsourced the letter, and their addresses sat in a platform they had never heard of. Le Figaro, citing Cyberattaque.org, reports that social housing providers, medical laboratory patients, and patients of Foch Hospital in Suresnes are among those affected. The Foch exposure is reported to come through a subcontractor, the debt-recovery firm Kiwee, and the reporting describes the hacker’s access to those letters as believed, not confirmed. We treat the downstream victim list the same way until the organizations confirm it.

The harm is quiet. A name and a postal address are not a password, but a list of who receives letters from which sender is raw material for phishing that looks like the real thing. The practical question for any organization that uses a mailing or document provider is whether it can list every processor that holds its customers’ addresses.

2. N-able: A researcher dropped a new N-central exploit chain after the fixes, and attackers used it five hours later

The headline: Researcher Sean Bales published more than 30 alleged zero-days in N-able’s N-central, which he named Able2Shell, and N-able took its hosted N-central environments offline as a precaution. Bales reportedly said he released them after N-able failed to patch defects he had reported in another product. N-able shipped fixes, and about two days later Bales posted another zero-day, 4EverAble2Shell. On November 12 he published proof-of-concept code for it and for Agent2Shell, a separate set of roughly a dozen flaws that could give attackers System privileges on managed Windows agents. SecurityWeek reports that Previdian recorded exploitation attempts within hours and that Lupovis and Huntback saw successful compromises about five hours after the PoC appeared. 4EverAble2Shell chains CVE-2026-90436, an Apache Axis flaw rated CVSS 9.8, with an unspecified flaw in N-central deployments that have the Legacy API enabled. Bales says it works against servers updated to 2026.4 HF2. The only payload seen so far is an XMRig cryptocurrency miner, and Huntback reports that 19 victims had applied N-able’s latest patches. N-able restricted the Legacy API and the Intune integration on November 13 and says it is working with affected customers.

What we’re actually watching: The patch did not settle it. Huntback warns that Hotfix 2 alone is not enough and that a hit server should be treated as fully compromised. A miner is the cheapest thing an intruder can drop, and root on a remote management server lets them drop anything else. The attacker’s wallet showed 21 mining rigs, nine of them in the United States, so the compromised servers are spread across several countries. Five hours separated a public proof of concept from live compromises.

N-central manages endpoints across many customer environments, which is why we flagged it in our August 3 roundup. Teams that run it, or that rely on a provider who does, should keep the Legacy API off, restrict the console to a trusted network, and ask the provider whether its instance was reachable. Angriffsflächenmanagement finds management consoles exposed to the internet before a scanner does.

3. Check Point: Attackers chained PHP2Shell with a privilege bug on gateways, then ran a wiper on a VMware host

The headline: Check Point Research published its analysis on November 12 of attacks on Check Point Security Gateways with a Mobile Access portal. The actor used PHP2Shell (CVE-2026-109618 and CVE-2026-109619) for initial access, then CVE-2026-123039, a privilege escalation in the DynamicCLI IPC interface, to reach root. It placed PHP webshells in the portal’s templates directory, deployed the TSGA backdoor and a scanner called gogo, and moved over SSH to a VMware ESXi host using the vpxuser account. On that host it ran a wiper, destroy_v6, that deletes files recursively, and left a politically themed note called DREAM_ANNEXATION.txt that rejects ransom negotiations. The first wave began on October 30 and hit Israeli organizations, with some destructive activity. A second wave from November 2 hit organizations in multiple countries and sectors, with no destructive activity observed. A security fix had been available for about a week before the first wave. SecurityWeek covered the findings.

What we’re actually watching: Check Point says the evidence does not let it assess the true nature of the operation with high confidence. The backdoor is a modified variant of PhiliKit, which ESET previously linked to the China-nexus group UNC5221, but the wiper, the Israel focus, and the Hebrew note do not fit that profile. Check Point lists three possibilities: one actor ran everything, the actor was only an initial access broker, or the destruction is a false flag or a smokescreen. We name no culprit, and a briefing built from this should not either.

The path matters more than the name. A firewall portal was the front door, root on the gateway was a map of the inside, and a hypervisor was where the damage landed. According to Rescana, CISA added CVE-2026-123039 to its Known Exploited Vulnerabilities catalog on November 13 with a federal deadline of November 16, and the fix ships in Check Point’s November Jumbo Hotfix Accumulator. Teams should patch the gateway, look for zz*.tpl files in the portal templates and the slot_CPconfigdaemon backdoor, and review vpxuser SSH logins on ESXi. Our October 26 roundup made the same point about exposed management interfaces on perimeter devices.

4. Microsoft: Patch Tuesday fixed about 1,090 CVEs, and the exploited one is rated only 6.5

The headline: Microsoft’s November Patch Tuesday fixed 1,093 CVEs by Tenable’s count und 1,094 by CyberScoop’s, with 63 rated Critical in Tenable’s tally. One flaw was exploited as a zero-day: CVE-2026-124194, an Outlook spoofing vulnerability scored CVSS 6.5 and rated Moderate. Per Tenable, an attacker who persuades a target to open a crafted Office file can obtain the target’s NTLMv2 hash and use it to impersonate them in follow-on attacks. CyberScoop describes the trigger as a specially crafted message file that makes Outlook connect to an attacker-controlled network location without the usual external-content warning, and says the bug was added to CISA’s Known Exploited Vulnerabilities catalog shortly after Microsoft’s advisory. No flaws were publicly known before release. The month is the second in a row to exceed the totals of all previous years, though it is smaller than October’s record 1,413.

What we’re actually watching: The score and the risk point in opposite directions. A 6.5 sits low in a patch queue of more than a thousand items, yet this is the one attackers were already using. Dustin Childs of TrendAI’s Zero Day Initiative cautions that fewer exploited zero-days does not mean less exploitation in the wild. He also singled out CVE-2026-116928, a Windows Search flaw reachable through Exchange Server that he called a zero-click code execution bug on a high-value target, and a Windows LDAP flaw rated 9.8 that he says could be destructive on domain controllers.

A response from Microsoft following the worldwide outage late last week.

Sort the queue by exploitation status first and severity score second. A stolen NTLMv2 hash is also exactly what turns up later in credential markets, so the patch does not end the exposure for anyone whose hash already left the building. Credential Intelligence detects stolen credentials across data breaches, paste sites, and dark web forums before attackers use them.

5. Operation Endgame: Authorities seized 235 servers and found 6.4 million victims’ credentials, with no arrests

The headline: Europol announced on November 12 a new phase of Operation Endgame against three malware families. Nine countries took part: Bulgaria, Canada, Denmark, Germany, Moldova, the Netherlands, Poland, the United Kingdom, and the United States. Across the operation, authorities seized 235 servers and 257 domains. CastleRAT, a remote access trojan known since 2025, accounted for 120 servers and 39 domains. Gootloader, a dropper spread through fake legal documents on hacked WordPress sites, accounted for 14 servers and 218 domains, and 681 infected websites were cleaned up. StealC, an infostealer with dropper functions, accounted for 102 servers, building on seizures from June 2026. Investigators identified credentials belonging to 6.4 million victims. According to the U.S. Department of Justice, Gootloader infected at least 25,000 computers, was used in ransomware attacks against at least 44 U.S. victims, and has caused about $28.8 million in losses since 2018. The Hacker News covered the takedown as well.

What we’re actually watching: Seizing servers ends the operators’ control, but the stolen data does not expire with them. Six million sets of credentials do not stop working because a command server went dark, and no arrests were reported, so the people behind these tools are free to rebuild. Victim data is being shared through Have I Been Pwned, the Shadowserver Foundation, and the Dutch National Cyber Security Centre so people can check their exposure. Authorities advised WordPress users to update credentials, enable multi-factor authentication, and patch vulnerabilities.

The pattern across all five stories

In every story this week, the official fix or action was not the end of the exposure: a notification months after the attack, a patch that did not hold, a Check Point fix that was available a week before the first wave and still missed by some, a low score on an exploited bug, a takedown that left 6.4 million credentials behind.

CybelAngel maps your external attack surface and finds the credentials and data already in circulation, so you can act before the next advisory. Our 2026 Supply Chain Cyber Risk Report shows how far that exposure reaches.

Über den Autor