|
| 1 | +# Suspicious Access to Sensitive Local User Documents Triage Guide |
| 2 | + |
| 3 | +## Rule Overview |
| 4 | + |
| 5 | +**Title:** Suspicious Access to Sensitive Local User Documents |
| 6 | +**Rule ID:** SENT-COLL-0005 |
| 7 | +**Severity:** Medium |
| 8 | +**Risk Score:** 57 |
| 9 | +**Tactic:** Collection |
| 10 | +**Technique:** T1005 - Data from Local System |
| 11 | +**Platform:** Microsoft Sentinel |
| 12 | +**Data Source:** DeviceFileEvents |
| 13 | +**Lifecycle:** Experimental |
| 14 | + |
| 15 | +## Purpose |
| 16 | + |
| 17 | +This detection identifies processes accessing potentially sensitive document types in common user data paths, which may indicate collection or staging from the local system. |
| 18 | + |
| 19 | +This matters because attackers often target user documents during collection activity in order to gather: |
| 20 | + |
| 21 | +- Business documents |
| 22 | +- PDFs |
| 23 | +- Financial spreadsheets |
| 24 | +- Presentations |
| 25 | +- CSV exports |
| 26 | +- Notes or text files containing sensitive information |
| 27 | + |
| 28 | +## Detection Logic Summary |
| 29 | + |
| 30 | +The rule reviews `DeviceFileEvents` for: |
| 31 | + |
| 32 | +- `FileCreated` |
| 33 | +- `FileModified` |
| 34 | +- `FileRead` |
| 35 | +- `FileAccessed` |
| 36 | + |
| 37 | +It focuses on file activity in common user data paths such as: |
| 38 | + |
| 39 | +- `\Users\` |
| 40 | +- `\Desktop\` |
| 41 | +- `\Documents\` |
| 42 | +- `\Downloads\` |
| 43 | + |
| 44 | +It includes document types such as: |
| 45 | + |
| 46 | +- `.docx` |
| 47 | +- `.pdf` |
| 48 | +- `.xls` |
| 49 | +- `.xlsx` |
| 50 | +- `.csv` |
| 51 | +- `.pptx` |
| 52 | +- `.txt` |
| 53 | + |
| 54 | +It excludes common expected processes such as: |
| 55 | + |
| 56 | +- `explorer.exe` |
| 57 | +- `SearchIndexer.exe` |
| 58 | +- `OneDrive.exe` |
| 59 | +- `MsMpEng.exe` |
| 60 | + |
| 61 | +The rule alerts when a process accesses at least 25 files across at least 3 distinct paths in a 15-minute window. |
| 62 | + |
| 63 | +## Likely Analyst Goal |
| 64 | + |
| 65 | +Determine whether the file access was: |
| 66 | + |
| 67 | +- Normal business activity |
| 68 | +- Backup, indexing, sync, or anti-malware behavior |
| 69 | +- Approved enterprise tooling |
| 70 | +- Suspicious local document collection or staging |
| 71 | + |
| 72 | +## Initial Triage Questions |
| 73 | + |
| 74 | +1. What process accessed the files? |
| 75 | +2. Which user was involved? |
| 76 | +3. Is this level of document access normal for the host or user? |
| 77 | +4. Are the accessed file types sensitive for that role? |
| 78 | +5. Did archive, upload, or transfer behavior follow? |
| 79 | + |
| 80 | +--- |
| 81 | + |
| 82 | +## Investigation Steps |
| 83 | + |
| 84 | +### 1. Review the Process and Account |
| 85 | + |
| 86 | +Inspect: |
| 87 | + |
| 88 | +- `InitiatingProcessFileName` |
| 89 | +- `InitiatingProcessCommandLine` |
| 90 | +- `InitiatingProcessAccountName` |
| 91 | + |
| 92 | +Determine whether the process is: |
| 93 | + |
| 94 | +- Common and trusted |
| 95 | +- Rare in the environment |
| 96 | +- Running from a suspicious path |
| 97 | +- Associated with scripting, automation, or portable tooling |
| 98 | + |
| 99 | +**Why this matters:** |
| 100 | +The process identity is one of the main clues for separating legitimate bulk access from suspicious collection. |
| 101 | + |
| 102 | +--- |
| 103 | + |
| 104 | +### 2. Review the Scope of File Access |
| 105 | + |
| 106 | +Evaluate: |
| 107 | + |
| 108 | +- Number of files accessed |
| 109 | +- Number of folders involved |
| 110 | +- Types of documents accessed |
| 111 | +- Whether the activity was concentrated in user data paths |
| 112 | + |
| 113 | +Determine whether the files appear to represent: |
| 114 | + |
| 115 | +- Normal working documents |
| 116 | +- Sensitive business content |
| 117 | +- Broad user data harvesting |
| 118 | + |
| 119 | +**Why this matters:** |
| 120 | +Wide file access across multiple user directories is consistent with collection or staging workflows. |
| 121 | + |
| 122 | +--- |
| 123 | + |
| 124 | +### 3. Determine Whether the Activity Is Expected |
| 125 | + |
| 126 | +Ask: |
| 127 | + |
| 128 | +- Is the process a backup, sync, or document management tool? |
| 129 | +- Is the user performing bulk review, migration, or archiving? |
| 130 | +- Is the activity associated with approved business software? |
| 131 | +- Does the role normally involve opening many documents? |
| 132 | + |
| 133 | +**Why this matters:** |
| 134 | +Legitimate tools can generate noisy document access patterns. |
| 135 | + |
| 136 | +--- |
| 137 | + |
| 138 | +### 4. Check for Follow-On Staging or Exfiltration |
| 139 | + |
| 140 | +Review whether the activity was followed by: |
| 141 | + |
| 142 | +- Archive creation |
| 143 | +- Cloud upload |
| 144 | +- Email attachment activity |
| 145 | +- Removable media use |
| 146 | +- Network transfer |
| 147 | +- File compression |
| 148 | +- Copies into temp or staging directories |
| 149 | + |
| 150 | +**Why this matters:** |
| 151 | +Document access followed by staging or transfer is much more suspicious than document access alone. |
| 152 | + |
| 153 | +--- |
| 154 | + |
| 155 | +### 5. Review the Execution Context |
| 156 | + |
| 157 | +Check: |
| 158 | + |
| 159 | +- Parent process |
| 160 | +- Signer information |
| 161 | +- Execution path |
| 162 | +- Whether the binary is known and prevalent |
| 163 | +- Whether it launched from a user-writable location |
| 164 | + |
| 165 | +Pay special attention to: |
| 166 | + |
| 167 | +- PowerShell |
| 168 | +- Rare executables |
| 169 | +- Temp directory launches |
| 170 | +- Unsigned binaries |
| 171 | + |
| 172 | +**Why this matters:** |
| 173 | +Execution context helps determine whether the activity is enterprise tooling or suspicious collection. |
| 174 | + |
| 175 | +--- |
| 176 | + |
| 177 | +### 6. Assess User and Device Context |
| 178 | + |
| 179 | +Review: |
| 180 | + |
| 181 | +- Whether the host is high value |
| 182 | +- Whether the account is privileged |
| 183 | +- Whether the device has recent suspicious alerts |
| 184 | +- Whether similar activity is normal on that system |
| 185 | + |
| 186 | +**Why this matters:** |
| 187 | +Collection activity on finance, HR, executive, or admin systems may raise priority. |
| 188 | + |
| 189 | +--- |
| 190 | + |
| 191 | +## Benign Explanations |
| 192 | + |
| 193 | +Common legitimate scenarios include: |
| 194 | + |
| 195 | +1. Backup, indexing, sync, or anti-malware activity |
| 196 | +2. Bulk document processing by IT or approved business tooling |
| 197 | +3. User-driven search, migration, or archival workflows |
| 198 | + |
| 199 | +--- |
| 200 | + |
| 201 | +## Suspicious Indicators |
| 202 | + |
| 203 | +Escalate concern when you observe: |
| 204 | + |
| 205 | +- Unknown or rare process touching many documents |
| 206 | +- Execution from temp or user profile paths |
| 207 | +- Sensitive file types across many directories |
| 208 | +- File access followed by compression or upload |
| 209 | +- Similar behavior on other hosts tied to the same account |
| 210 | +- Additional collection or exfiltration alerts nearby |
| 211 | + |
| 212 | +--- |
| 213 | + |
| 214 | +## Triage Decision |
| 215 | + |
| 216 | +### Close as Benign / False Positive |
| 217 | + |
| 218 | +Close as benign when: |
| 219 | + |
| 220 | +- The process is an approved business or enterprise tool |
| 221 | +- The account and host commonly perform bulk file operations |
| 222 | +- No staging or exfiltration activity is observed |
| 223 | + |
| 224 | +### Escalate as Suspicious |
| 225 | + |
| 226 | +Escalate when: |
| 227 | + |
| 228 | +- The process is uncommon or poorly understood |
| 229 | +- The file access volume is unusual for the user or host |
| 230 | +- Follow-on staging or transfer activity is present |
| 231 | + |
| 232 | +### Escalate as Likely Malicious |
| 233 | + |
| 234 | +Escalate as likely malicious when: |
| 235 | + |
| 236 | +- Bulk document access is tied to suspicious tooling |
| 237 | +- Archive, transfer, or exfiltration clearly follows |
| 238 | +- The host shows broader compromise evidence |
| 239 | + |
| 240 | +--- |
| 241 | + |
| 242 | +## Response Actions |
| 243 | + |
| 244 | +Depending on findings, consider: |
| 245 | + |
| 246 | +- Isolating the host if large-scale collection is suspected |
| 247 | +- Collecting the process binary and hash |
| 248 | +- Hunting for the same process across the environment |
| 249 | +- Reviewing cloud, email, and removable media activity |
| 250 | +- Escalating to incident response if staging or exfiltration is confirmed |
| 251 | + |
| 252 | +--- |
| 253 | + |
| 254 | +## Example Analyst Notes Template |
| 255 | + |
| 256 | +### Analyst Summary |
| 257 | + |
| 258 | +Alert fired for suspicious access to sensitive local user documents, potentially indicating collection or staging from the local system. |
| 259 | + |
| 260 | +### Key Findings |
| 261 | + |
| 262 | +- **Affected device:** |
| 263 | +- **Affected user:** |
| 264 | +- **Process:** |
| 265 | +- **Command line:** |
| 266 | +- **Files accessed:** |
| 267 | +- **Folders involved:** |
| 268 | +- **Expected business purpose:** |
| 269 | +- **Nearby archive or upload activity:** |
| 270 | +- **Final assessment:** |
| 271 | + |
| 272 | +### Recommended Disposition |
| 273 | + |
| 274 | +- Benign / False Positive |
| 275 | +- Suspicious - Needs Deeper Investigation |
| 276 | +- Confirmed Malicious |
0 commit comments