Skip to main content
Mallory
Malware4 associated actors2 associated CVEs

Ghost

Tracked since Nov 25, 2025 · Updated today
Trending this weekNot trendingOutside this week’s top 100
Mention activity · 30 days4 mentions · peak 2 in a dayToday
Vulns2Associated
Actors4Associated
ATT&CK6Techniques
IOCs9Tracked
Type
Backdoor
Capabilities
ReconnaissanceScanningBrute forceInitial accessCredential theftExfiltrationPost exploitationPersistenceDefense evasion
Delivery
Seo poisoningLure
Platforms
Windows

Ghost is an AI-enabled offensive agent in the BlackHatSect0r toolchain, using the runtime name GHOSTHUNTERS and the Hermes agent framework. It translates natural-language operator requests into scanners, exploit code, data-processing utilities, exfiltration scripts and implants. Its configuration incorporates anti-refusal prompts and a local proxy that retries rejected model requests with escalating prompts.

The French-speaking operator BabidiGhost used Ghost during November 6–14, 2026, to target more than 50 organizations and datasets, primarily in Belgium, France and Switzerland. Targets included municipalities, banks, insurers, payment providers, cryptocurrency businesses and betting platforms. Successful intrusions relied on known vulnerabilities, exposed secrets, weak passwords and unauthenticated services rather than zero-day exploitation. At an insurer, exploitation of CVE-2023-38646 in Metabase enabled server command execution, theft of cloud database credentials and export of customer data. Other activity included attack-surface mapping, account enumeration, authentication brute-force attempts and theft of municipal archives from an exposed storage-management interface.

Ghost also supported deployment of PowerShell and C# implants. Persistence and evasion mechanisms included database-trigger execution, scheduled tasks, Registry autorun entries, Windows Update service impersonation and Microsoft Defender exclusions. Revocation of its model API key stopped agent activity on November 14, although previously deployed implants continued communicating into the following day.

Is this in your environment?See whether this family's indicators match your telemetry and which of your assets are exposed to the CVEs associated with it.
Check your exposure
Associated CVEs

Associated vulnerabilities

2 CVEs Mallory has correlated with this family across public research and vendor advisories. Each card links to the full Mallory page for that vulnerability.

CVE-2023-38646via socradar blogPre-authentication Remote Code Execution in MetabaseIts identity file introduces it as Ghost, a “BlackHatSect0r Agent,” and names its runtime GHOSTHUNTERS , the same signature documented in Part I.View vulnerability →Source ↗
CVE-2018-13379via reliaquest com threat huntingExploited in the wildUnauthenticated Path Traversal in Fortinet FortiOS and FortiProxy SSL VPNExploiting CVE-2018-13379 gives adversaries direct, unauthenticated access to sensitive system files, including the jackpot sslvpn_websession. This plaintext file often contains usernames and passwords, giving attackers immediate access to credentials. | On February 19, 2025, CISA and the FBI issued an advisory warning about “Ghost” ransomware, which exploits this vulnerability using publicly available code.View vulnerability →Source ↗
Threat actors

Associated threat actors

4 distinct threat actors Mallory has associated with this family across public reporting. Open in Mallory to see the full evidence chain and overlapping campaigns.

View more details →
BlackHatSect0r

Its identity file introduces it as Ghost, a “BlackHatSect0r Agent,” and names its runtime GHOSTHUNTERS , the same signature documented in Part I.

via socradar blogsocradar.io
BabidiGhost

Its identity file introduces it as Ghost, a “BlackHatSect0r Agent,” and names its runtime GHOSTHUNTERS , the same signature documented in Part I.

via socradar blogsocradar.io
Utg Q 015

UTG-Q-015 generally use Xnote, Ghost, Vshell and other backdoors to control the target linux servers.

via qianxin ti centerti.qianxin.com
Threat Group-3390

APT27 (aka Lucky Mouse, Emissary Panda, Iron Tiger, ZipToken, Group 35, TEMP.Hippo, TG 3390, Bronze Union) ... Examples of associated tools: Ghost, ASPXSpy, ZxShell RAT, HyperBro, PlugX RAT, Windows Credential Editor, FoundCore, China Chopper...

via cert europacert.europa.eu
MITRE ATT&CK

Techniques & procedures

6 distinct techniques documented for this family, organized by ATT&CK tactic.

Reconnaissance

1 technique
T1595.002Vulnerability ScanningEvidence1

Certificate and DNS harvesting, path scanning, account validation

Execution

2 techniques
T1059.006PythonEvidence1

PowerShell beacons; Python tasking from the agent

T1204.002Malicious FileEvidence1

“When a user ran the disguised installation file, shellcode ... was triggered, and the malicious payload was executed.”

Persistence

1 technique
T1543.003Windows ServiceEvidence1

“In the latest variant, the malicious file was created in the C:\msys64 Path, registered as a Service, and executed.”

Privilege Escalation

1 technique
T1543.003Windows ServiceEvidence1

“In the latest variant, the malicious file was created in the C:\msys64 Path, registered as a Service, and executed.”

Stealth

1 technique
T1036MasqueradingEvidence1

“[A] malicious installer disguised as a KakaoTalk installer” and “Malicious file names and creation paths consisted of random strings.”

Impact

1 technique
T1486Data Encrypted for ImpactEvidence1

The attacks led to the threat actors encrypting files, and in one case, even corrupting a computer used to control the SCADA industrial equipment deployed inside the treatment plant.

Indicators of compromise

Observables

9 indicators attributed across vendor reports, sandbox runs, and researcher write-ups. Values are masked here and available in full in Mallory.

View all in app →
Indicator values are available in Mallory. Pivot on every hash, domain, URL, and IP, or pipe them straight into your SIEM.View all in Mallory
Activity feed

Recent activity

21 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.

View more in app →
malware newsNews · Nov 20, 2026BabidiGhost: An AI-Augmented Cybercrime Operation - Malware News - Malware Analysis, News and IndicatorsA jailbroken Hermes-based AI agent that translates plain-language orders into scanning, exploitation, credential harvesting, database export and exfiltration scripts. Its anti-refusal configuration and retry proxy enabled sustained offensive activity despite limited operator skill and framework failures. BabidiGhost used it against more than 50 organizations and datasets; revocation of its model API key stopped agent activity on November 14, 2026, but deployed implants continued checking in.Read more ↗socradar blogNews · Nov 20, 2026BabidiGhost: An AI-Augmented Cybercrime OperationA jailbroken agent built on the Hermes framework that converts plain-language orders into scanning, exploitation, credential harvesting, data theft and implant-deployment activity. Its anti-refusal instructions and escalation mechanism suppress model refusals. BabidiGhost used it against European organizations, exploiting known weaknesses including CVE-2023-38646. Revocation of its model API key stopped agent activity on November 14, 2026, but deployed implants continued operating.Read more ↗cryptika blogNews · Oct 21, 2026Hackers Use Claude AI to Plan Ransomware Attacks, Then Accidentally Expose Their Secrets | Cryptika CybersecurityCommand-and-control software present in the exposed operators’ toolkit. The indicators identify a Ghost panel and Ghost delivery infrastructure, but the content does not describe its implementation or specific actions on victim systems.Read more ↗cyber security newsNews · Oct 21, 2026Hackers Use Claude AI to Plan Ransomware Attacks, Then Accidentally Expose Their SecretsCommand-and-control software found in the ransomware operators’ toolkit. The listed infrastructure included a Ghost panel and a server associated with Ghost and payload delivery.Read more ↗
+15 more in the timeline. Sign up to see them
Is this in your environment?See whether this family's indicators match your telemetry and which of your assets are exposed to the CVEs associated with it.
Check your exposure
Mallory platform

The version that knows your environment.

This page is what’s public. Mallory adds the parts that aren’t: which of your assets match these IOCs, which detections are missing, which campaigns to expect next, and what to do in the next 30 minutes.
IOC matching9Match every observed IP, domain, and hash against your live telemetry.
Associated threat actors4Groups linked to this family, with evidence pinned to each claim.
Associated vulnerabilities2CVEs linked to this family, with the reporting behind each link.
Detection signaturesYARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
MITRE ATT&CK mapping6Every documented technique, ranked by evidence weight.
Researcher chatterReddit, Mastodon, and CTI community discussion around this family.