Valid Accounts: Cloud Accounts
121 detection findings mapped across 3 cloud providers; 9 CloudSigma rules.
View coverage →Last reviewed:
32 ATT&CK techniques in this batch, grouped by tactic. Click a tactic card to jump to its techniques. New techniques graduate as their CloudSigma rule sets land.
Jump to a tactic:
How adversaries get into the environment.
4 techniques TA0002How adversaries run their code once inside.
1 techniques TA0003How adversaries keep their foothold across reboots and credential rotations.
3 techniques TA0004How adversaries gain higher privileges than they were given.
3 techniques TA0006How adversaries steal credentials, account names and passwords.
3 techniques TA0007How adversaries learn what they have access to.
5 techniques TA0009How adversaries gather data of interest before exfiltration.
2 techniques TA0040How adversaries disrupt or destroy systems and data.
5 techniques TA0042How adversaries acquire infrastructure they will use later.
1 techniques TA0112How adversaries disable or degrade the defences that would have spotted them.
5 techniquesCloud scope Cloud-account abuse and exploitation of cloud-hosted public-facing applications. Phishing, removable media, drive-by compromise, supply-chain attacks on endpoint software, and trusted-relationship abuse via on-prem networks are endpoint and network vectors and are out of scope for cloud-audit-log detection.
121 detection findings mapped across 3 cloud providers; 9 CloudSigma rules.
View coverage →75 detection findings mapped across 3 cloud providers; 11 CloudSigma rules.
View coverage →60 detection findings mapped across 2 cloud providers; 11 CloudSigma rules.
View coverage →Descriptive entry: research-linked technique without victim-side cloud-audit-log signal of its own.
View coverage →Cloud scope Cloud-API misuse, serverless / container execution paths, and cloud-shell or runtime-environment abuse. Endpoint command and scripting interpreters (PowerShell, bash, AppleScript), user execution of malware, and exploitation for client execution are EDR territory and out of scope here.
18 detection findings mapped across 3 cloud providers; 6 CloudSigma rules.
View coverage →Cloud scope Cloud-account manipulation, additional cloud credentials, lambda / function injection, IAM role and policy abuse. Boot or logon persistence on endpoints (registry run keys, scheduled tasks, browser extensions, BIOS implants) is EDR territory.
27 detection findings mapped across 3 cloud providers; 3 CloudSigma rules.
View coverage →20 detection findings mapped across 3 cloud providers; 13 CloudSigma rules.
View coverage →19 detection findings mapped across 2 cloud providers; 4 CloudSigma rules.
View coverage →Cloud scope Cloud-IAM privilege escalation: role assumption, policy modification, cross-account chains, and over-permissive trust relationships. Endpoint UAC bypass, exploit-for-privilege-escalation, and process-injection techniques are out of scope.
18 detection findings mapped across 2 cloud providers; 1 CloudSigma rules.
View coverage →17 detection findings mapped across 3 cloud providers; 4 CloudSigma rules.
View coverage →2 detection findings mapped across 2 cloud providers; 6 CloudSigma rules.
View coverage →Cloud scope Cloud-credential theft via metadata-service abuse, token theft, and secrets-store mining (Secrets Manager, Key Vault, GCP Secret Manager). OS credential dumping (Mimikatz against LSASS, reads of /etc/shadow, browser-stored password extraction) is endpoint territory and out of scope.
57 detection findings mapped across 2 cloud providers; 3 CloudSigma rules.
View coverage →42 detection findings mapped across 3 cloud providers; 4 CloudSigma rules.
View coverage →10 detection findings mapped across 3 cloud providers; 6 CloudSigma rules.
View coverage →Cloud scope Cloud-resource enumeration via List* / Describe* / Get* API calls. Process discovery, file and directory discovery, account discovery on endpoints, and domain-trust enumeration require EDR or AD-side telemetry and are out of scope here.
60 detection findings mapped across 3 cloud providers; 3 CloudSigma rules.
View coverage →12 detection findings mapped across 2 cloud providers; 3 CloudSigma rules.
View coverage →10 detection findings mapped across 2 cloud providers; 3 CloudSigma rules.
View coverage →1 detection findings mapped across 1 cloud providers; 3 CloudSigma rules.
View coverage →1 detection findings mapped across 1 cloud providers; 3 CloudSigma rules.
View coverage →Cloud scope Cloud-storage and cloud-API data collection: S3 / Blob / GCS reads, snapshot exfil staging, repository collection from cloud-hosted code platforms. Endpoint clipboard, screen capture, audio capture, and local email collection are out of scope.
108 detection findings mapped across 3 cloud providers; 3 CloudSigma rules.
View coverage →1 detection findings mapped across 1 cloud providers; 3 CloudSigma rules.
View coverage →Cloud scope Cloud-resource destruction or denial (S3 object deletion, RDS instance drop, EC2 stop, snapshot deletion), data encryption for impact via cloud-managed keys, and DoS against cloud endpoints. Endpoint ransomware encrypting local disks and disk-wiping malware are EDR territory.
60 detection findings mapped across 3 cloud providers; 3 CloudSigma rules.
View coverage →48 detection findings mapped across 2 cloud providers; 3 CloudSigma rules.
View coverage →34 detection findings mapped across 3 cloud providers; 3 CloudSigma rules.
View coverage →25 detection findings mapped across 2 cloud providers; 3 CloudSigma rules.
View coverage →18 detection findings mapped across 3 cloud providers; 3 CloudSigma rules.
View coverage →Cloud scope Limited cloud signal: adversaries rarely build infrastructure inside victim cloud accounts pre-attack. Where they do (e.g. spinning up compute for cryptomining), the activity surfaces under Impact (T1496) and is detected via cloud audit logs there.
Descriptive entry: research-linked technique without victim-side cloud-audit-log signal of its own.
View coverage →Cloud scope Disabling cloud logging or cloud firewalls, modifying cloud-control configurations to hide activity, and tampering with cloud authentication or policy controls. Endpoint defence tampering (host firewall edits, AV or EDR disable, on-host artefact removal) is endpoint territory and out of scope.
66 detection findings mapped across 3 cloud providers; 3 CloudSigma rules.
View coverage →60 detection findings mapped across 3 cloud providers; 3 CloudSigma rules.
View coverage →1 detection findings mapped across 1 cloud providers; 3 CloudSigma rules.
View coverage →Descriptive entry: research-linked technique without victim-side cloud-audit-log signal of its own.
View coverage →Descriptive entry: research-linked technique without victim-side cloud-audit-log signal of its own.
View coverage →