Identify and Investigate Phishing Attacks with KQL
To detect potential phishing emails sent to users using KQL in Microsoft Sentinel or Microsoft Defender for Office 365, you can leverage the EmailEvents
table. Below are some examples that can be used to identify suspicious emails:
1. Basic Query to Detect Emails with Suspicious Attachments
Phishing emails often contain malicious attachments. Look for emails with suspicious file extensions.
2. Detect Emails with Suspicious URLs
Phishing emails often contain links to malicious websites. Check for emails containing URLs and filters based on known suspicious domains.
3. Detect Emails from External Senders with High-Risk Indicators
Identify emails from external senders with high-risk indicators such as spoofed display names or mismatched sender domains.
4. Detect Emails with Spoofed Display Names
Attackers often spoof display names to make emails appear as though it's from a trusted source. Look for these cases.
5. Detect Emails with High Spam Confidence Level (SCL)
Microsoft assigns a Spam Confidence Level (SCL) to emails. A higher SCL indicates a higher likelihood of spam or phishing.
6. Detect Emails with Phishing Keywords in the Subject
Phishing emails often use specific keywords in the subject line to lure victims. Look for common phishing-related keywords.
7. Detect Emails with Mismatched Sender Domains
Identify emails where the sender's domain does not match the domain in the email address, which could indicate spoofing.
8. Detect Emails with High Volume Sent to Multiple Users
Phishing campaigns often target multiple users at once. Check for emails sent to a large number of recipients.
9. Detect Emails with Malware Detected
Identify emails where malware was detected by Microsoft Defender for Office 365.
10. Detect Emails with Anomalous Sender Behavior
Look for emails from senders who have never communicated with your organisation before, which could indicate a phishing attempt.
Notes:
Replace
"yourcompany.com"
with your actual company domain.You can adjust the thresholds (e.g.,
SpamConfidenceLevel
,RecipientCount
) based on your organisation's risk tolerance.These queries can be further customised based on your organisation's specific needs and threat intelligence.
Last updated