Skip to content

Commit da588e6

Browse files
authored
Update graph-mail-search-keyword-burst.md
1 parent 3d295b6 commit da588e6

1 file changed

Lines changed: 264 additions & 0 deletions

File tree

Lines changed: 264 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,264 @@
1+
# Microsoft Graph Mail Access Burst Triage Guide
2+
3+
## Rule Overview
4+
5+
**Title:** Microsoft Graph Mail Access Burst
6+
**Rule ID:** SENT-COLL-0004
7+
**Severity:** Medium
8+
**Risk Score:** 70
9+
**Tactic:** Collection
10+
**Technique:** T1114 - Email Collection
11+
**Platform:** Microsoft Sentinel
12+
**Data Source:** CloudAppEvents
13+
**Lifecycle:** Experimental
14+
15+
## Purpose
16+
17+
This detection identifies bursts of Microsoft Graph mail search or mail access activity that may indicate post-compromise mailbox collection or reconnaissance.
18+
19+
This matters because attackers with cloud access may use Microsoft Graph to:
20+
21+
- Search mailbox contents
22+
- Read emails at scale
23+
- Collect sensitive communications
24+
- Enumerate high-value targets
25+
- Prepare for theft, extortion, or further compromise
26+
27+
## Detection Logic Summary
28+
29+
The rule reviews `CloudAppEvents` where:
30+
31+
- `Application == "Microsoft Graph"`
32+
- `ActionType` includes:
33+
- `SearchQueryPerformed`
34+
- `MailItemsAccessed`
35+
- `MessageBind`
36+
37+
It summarizes activity by:
38+
39+
- 30-minute window
40+
- user
41+
- application
42+
43+
The rule alerts when the action count reaches 20 or more within the window.
44+
45+
## Likely Analyst Goal
46+
47+
Determine whether the burst of Graph mail access was:
48+
49+
- Approved admin, migration, journaling, or eDiscovery activity
50+
- Legitimate application integration behavior
51+
- Suspicious mailbox reconnaissance or collection after identity compromise
52+
53+
## Initial Triage Questions
54+
55+
1. Which user or application performed the mail access?
56+
2. Is this volume of Graph mail activity normal?
57+
3. Was the activity tied to a known tenant application or integration?
58+
4. Were there recent risky sign-ins, device code sign-ins, or consent events?
59+
5. Did the activity target high-value mailboxes or lead to forwarding, export, or download behavior?
60+
61+
---
62+
63+
## Investigation Steps
64+
65+
### 1. Validate the User or Application Context
66+
67+
Review:
68+
69+
- User identity
70+
- Account type
71+
- Associated application
72+
- Source IP addresses
73+
74+
Determine whether the activity came from:
75+
76+
- A human user
77+
- A service principal
78+
- An approved integration
79+
- An unknown or suspicious application workflow
80+
81+
**Why this matters:**
82+
Graph access by approved enterprise applications is common, but unexpected users or apps can indicate abuse.
83+
84+
---
85+
86+
### 2. Review the Volume and Timing
87+
88+
Assess:
89+
90+
- Number of actions
91+
- Time window
92+
- Types of actions observed
93+
- Whether the burst is isolated or recurring
94+
95+
Determine whether the pattern suggests:
96+
97+
- Routine application polling
98+
- Large-scale mail review
99+
- Sudden collection after sign-in
100+
- Unusual mailbox targeting
101+
102+
**Why this matters:**
103+
A concentrated burst of mail access can indicate targeted collection or reconnaissance.
104+
105+
---
106+
107+
### 3. Review Authentication and Identity Signals
108+
109+
Check for nearby:
110+
111+
- Device code sign-ins
112+
- Risky sign-ins
113+
- Unfamiliar IP addresses
114+
- Impossible travel
115+
- MFA changes
116+
- OAuth abuse indicators
117+
- Consent grants
118+
119+
**Why this matters:**
120+
Mailbox collection often follows identity compromise or OAuth abuse.
121+
122+
---
123+
124+
### 4. Determine Whether the Activity Is Approved
125+
126+
Validate whether the activity aligns to:
127+
128+
- Migration tools
129+
- eDiscovery workflows
130+
- Journaling solutions
131+
- Mail security products
132+
- Approved enterprise applications
133+
- Known automation
134+
135+
**Why this matters:**
136+
Some applications legitimately access mailbox data at scale.
137+
138+
---
139+
140+
### 5. Check for High-Value or Targeted Mailbox Access
141+
142+
Determine whether the activity involved:
143+
144+
- Executives
145+
- Finance
146+
- HR
147+
- Legal
148+
- Admins
149+
- Sensitive shared mailboxes
150+
151+
Also assess whether the user accessed:
152+
153+
- Their own mailbox only
154+
- Multiple mailboxes
155+
- Unexpected high-value targets
156+
157+
**Why this matters:**
158+
Targeting sensitive mailboxes can indicate focused intelligence gathering.
159+
160+
---
161+
162+
### 6. Review for Follow-On Collection or Exfiltration
163+
164+
Look for nearby indicators such as:
165+
166+
- Mail export
167+
- Forwarding rule creation
168+
- Inbox rule changes
169+
- Download behavior
170+
- Additional Graph enumeration
171+
- SharePoint or OneDrive access bursts
172+
173+
**Why this matters:**
174+
Mail access becomes more serious when followed by export, forwarding, or broader cloud collection.
175+
176+
---
177+
178+
## Benign Explanations
179+
180+
Common legitimate scenarios include:
181+
182+
1. Migration tools
183+
2. eDiscovery, journaling, or approved admin search workflows
184+
3. Application integrations that legitimately access mail at scale
185+
186+
---
187+
188+
## Suspicious Indicators
189+
190+
Escalate concern when you observe:
191+
192+
- Graph mail access by an unusual user or app
193+
- Device code or risky sign-ins nearby
194+
- New or suspicious consent grants
195+
- Access from unfamiliar IP addresses
196+
- Multiple sensitive mailboxes accessed
197+
- Follow-on export, forwarding, or download activity
198+
199+
---
200+
201+
## Triage Decision
202+
203+
### Close as Benign / False Positive
204+
205+
Close as benign when:
206+
207+
- The user or application is approved
208+
- Activity matches known admin or business workflows
209+
- No suspicious sign-in or follow-on behavior is observed
210+
211+
### Escalate as Suspicious
212+
213+
Escalate when:
214+
215+
- The access burst is unusual for the user or app
216+
- Identity anomalies are present
217+
- High-value mailboxes were touched
218+
- Follow-on collection behavior is suspected
219+
220+
### Escalate as Likely Malicious
221+
222+
Escalate as likely malicious when:
223+
224+
- Evidence supports OAuth abuse or compromised credentials
225+
- Sensitive mailbox access is unexplained
226+
- Export, forwarding, or additional collection is confirmed
227+
228+
---
229+
230+
## Response Actions
231+
232+
Depending on findings, consider:
233+
234+
- Disabling or restricting the affected account or application
235+
- Revoking tokens or OAuth grants
236+
- Reviewing mailbox audit logs
237+
- Investigating related cloud collection activity
238+
- Escalating to incident response for suspected mailbox compromise
239+
240+
---
241+
242+
## Example Analyst Notes Template
243+
244+
### Analyst Summary
245+
246+
Alert fired for a burst of Microsoft Graph mail access activity, potentially indicating mailbox reconnaissance or collection.
247+
248+
### Key Findings
249+
250+
- **Affected user or application:**
251+
- **Source IPs:**
252+
- **Action volume:**
253+
- **Action types:**
254+
- **Expected business purpose:**
255+
- **Risky sign-in or consent activity:**
256+
- **High-value mailbox access:**
257+
- **Follow-on export or forwarding behavior:**
258+
- **Final assessment:**
259+
260+
### Recommended Disposition
261+
262+
- Benign / False Positive
263+
- Suspicious - Needs Deeper Investigation
264+
- Confirmed Malicious

0 commit comments

Comments
 (0)