↳ GitHub sourceAnalytics ruleMedium
Suspicious access of BEC related documents in AWS S3 buckets
Description
'This query looks for users with suspicious spikes in the number of files accessed that relate to topics commonly accessed as part of Business Email Compromise (BEC) attacks.
The query looks for access to files in AWS S3 storage that relate to topics such as invoices or payments, and then looks for users accessing these files in significantly higher numbers than in the previous 14 days. Incidents raised by this analytic should be investigated to see if the user accessing these files should be accessing them, and if the volume they accessed them at was related to a legitimate business need.
This query contains thresholds to reduce the chance of false positives, these can be adjusted to suit individual environments. In addition false positives could be generated by legitimate, scheduled actions that occur less often than every 14 days, additional exclusions can be added for these actions on username or IP address entities.'
- Rule type
- Scheduled
- Version
- 1.0.4
- Query frequency
- 1d
- Query period
- 14d
- Trigger
- gt 0
Declared MITRE coverage
Declared sources
Metadata from the source file. No dependencies inferred from KQL.
Connectors
Data types
KQL query
Original query, unchanged.
let BEC_Keywords = dynamic([ 'invoice','payment','paycheck','transfer','bank statement','bank details','closing','funds','bank account','account details','remittance','purchase','deposit',"PO#","Zahlung","Rechnung","Paiement", "virement bancaire","Bankuberweisung",'hacked','phishing']);
// Adjust this threshold based on your environment
let sensitivity = 2.5;
let Events = materialize(AWSCloudTrail
| where TimeGenerated between (ago(14d)..ago(0d))
| where UserIdentityAccountId != "anonymous"
| where EventSource startswith "s3."
| where EventName =~ "GetObject"
| extend FilePath = tostring(parse_json(RequestParameters).key)
| where FilePath has_any(BEC_Keywords)
);
Events
| summarize dcount(FilePath) by UserIdentityPrincipalid, bin(startofday(TimeGenerated), 1d)
| summarize CountOfDocs = make_list(dcount_FilePath, 10000), TimeStamp = make_list(TimeGenerated, 10000) by UserIdentityPrincipalid
| extend (Anomalies, Score, Baseline) = series_decompose_anomalies(CountOfDocs, sensitivity, -1, 'linefit')
| mv-expand CountOfDocs to typeof(double), TimeStamp to typeof(datetime), Anomalies to typeof(double),Score to typeof(double), Baseline to typeof(long)
| where Anomalies > 0
| project TimeStamp, CountOfDocs, Baseline, Score, Anomalies, UserIdentityPrincipalid
| join kind=inner(Events | extend TimeStamp = startofday(TimeGenerated)) on TimeStamp, UserIdentityPrincipalid
| extend Name = iif(UserIdentityUserName contains "@", split(UserIdentityUserName, "@")[0], UserIdentityUserName)
| extend UPNSuffix = iif(UserIdentityUserName contains "@", split(UserIdentityUserName, "@")[1], "")
| project-reorder TimeGenerated, UserIdentityType, UserIdentityPrincipalid, UserIdentityUserName, FilePath, EventName, UserAgent, SourceIpAddress, CountOfDocs, Baseline, Score
Declared entities
Related content
Links established from declared identifiers and solution manifests.
Source provenance
GitHubDisplayed values come from files in Azure/Azure-Sentinel. They describe the published template, not your workspace configuration.
- Commit
9800e51↗- Source identifier
f3e2d35f-1202-4215-995c-4654ef07d1d8
GSTEP / CATALOG TRACKING
Added to catalog : 16 Sept 2026 · 05:49 UTC
Last change observed : 16 Sept 2026 · 05:49 UTC