Lesson 37 — LSASS Was Accessed by an Unexpected Process
The endpoint was already suspicious.
PowerShell had launched. A file appeared and executed. The process contacted an external IP and created persistence.
Then another event raised the stakes:
an unexpected process accessed LSASS.
LSASS is involved in Windows authentication and can contain valuable credential material.
Legitimate security and system processes may interact with it, but unexpected access during an active attack chain demands careful investigation.
Credential access changed the risk
The analyst must identify the accessing process, establish why it touched LSASS, connect it to the existing execution chain and look for signs of credential use afterwards.
Case briefing
Investigation objective
Determine whether the LSASS access is expected, identify the process and ancestry involved, validate security or administrative explanations and assess whether credential-access behaviour may have enabled later movement.
Investigator's rule
Credential-access indicators need context. The same LSASS interaction can be legitimate in one environment and highly suspicious in another.
Stage 1 — understand why LSASS matters
| Question | Why it matters |
|---|---|
| Which process accessed LSASS? | Identifies the actor inside the endpoint timeline. |
| Is the process expected to interact with LSASS? | Security tools and operating-system components may have legitimate reasons. |
| What launched the process? | Process ancestry may connect the access to the earlier attack chain. |
| What happened immediately afterwards? | New logons or movement can strengthen the credential-access hypothesis. |
| Is the behaviour common in the environment? | Prevalence helps distinguish baseline activity from anomaly. |
LSASS is not inherently suspicious
Windows and security software interact with LSASS legitimately. The investigation must identify why this specific process accessed it.
The existing chain matters
An unknown process touching LSASS after suspicious execution, external communication and persistence deserves far more scrutiny than the same event in isolation.
Stage 2 — identify suspicious LSASS-related process activity
let TargetDevice = "WS-FIN-044";
DeviceProcessEvents
| where Timestamp > ago(24h)
| where DeviceName =~ TargetDevice
| where FileName =~ "update-check.exe"
or ProcessCommandLine has "lsass"
| project Timestamp,
AccountName,
FileName,
FolderPath,
SHA1,
ProcessCommandLine,
InitiatingProcessFileName,
InitiatingProcessCommandLine
| order by Timestamp asc
Start with the known suspicious process
The earlier lessons already established update-check.exe as part of the attack chain. Use that knowledge to reduce noise and focus the investigation.
Command-line evidence may be indirect
Not every LSASS-access event exposes intent directly in the process command line. Use the telemetry that is available and avoid inventing details that are not present.
Stage 3 — look for Defender endpoint events around the access
let StartTime = datetime(2026-08-25 10:15:30);
let EndTime = datetime(2026-08-25 10:17:00);
DeviceEvents
| where DeviceName =~ "WS-FIN-044"
| where Timestamp between (StartTime .. EndTime)
| project Timestamp,
ActionType,
AccountName,
InitiatingProcessFileName,
InitiatingProcessCommandLine,
AdditionalFields
| order by Timestamp asc
Read the event detail, not just the action name
Relevant Defender events can vary by detection and platform context. Preserve ActionType and AdditionalFields so the investigation remains grounded in actual telemetry.
Correlate the initiating process
If the same executable associated with file creation, network activity and persistence also appears in the LSASS-related event, the attack chain becomes substantially stronger.
Stage 4 — reconstruct the attack sequence
The hypothesis has changed
Earlier lessons asked whether the endpoint was compromised. At this point the SOC must also consider whether credentials may now be exposed.
Credential access can expand scope
If valid credentials were obtained, the incident may no longer be limited to one endpoint. Identity and lateral-movement hunting become necessary.
Stage 5 — check whether the process is normal in the environment
let TargetHash = "9D0B...E41C";
DeviceProcessEvents
| where Timestamp > ago(30d)
| where SHA1 == TargetHash
| summarize
Executions=count(),
Devices=dcount(DeviceName),
FirstSeen=min(Timestamp),
LastSeen=max(Timestamp)
by SHA1, FileName
Low prevalence adds weight
A process seen only on the compromised device deserves more scrutiny, especially when it already sits inside a suspicious attack timeline.
High prevalence can still be suspicious
Trusted or common binaries can be abused. Prevalence helps with context but does not replace process ancestry and behaviour.
Stage 6 — test legitimate explanations
| Possible explanation | What to validate |
|---|---|
| Endpoint security software | Vendor process, signed binary, expected deployment and documented LSASS interaction. |
| Administrative troubleshooting | Known administrator, approved tool, change record and expected timing. |
| System component | Expected Windows path, signature and normal parent process. |
| Credential-access behaviour | Suspicious ancestry, low prevalence, attack-chain timing and unexplained LSASS interaction. |
Digital signatures are useful context
A valid signature can support a legitimate explanation, but signed software can still be abused. Treat signature status as one part of the evidence.
Ask the endpoint team
Security tooling can generate low-level process access that looks unusual. Validate known EDR, antivirus and credential-management software before escalating solely on LSASS access.
Stage 7 — look for follow-on logon activity
let TargetAccount = "alex.wilson";
DeviceLogonEvents
| where Timestamp > ago(24h)
| where AccountName =~ TargetAccount
| project Timestamp,
DeviceName,
LogonType,
ActionType,
AccountDomain,
AccountName,
RemoteIP,
RemoteDeviceName
| order by Timestamp asc
Do not assume credentials were successfully extracted
Unexpected LSASS access raises the hypothesis. Follow-on authentication evidence is needed before claiming stolen credentials were used.
Look for the next device
If the same account appears on another endpoint shortly afterwards, that new device becomes a high-priority pivot for possible lateral movement.
Stage 8 — make the SOC decision
Write the investigation finding
Lesson 37 key takeaways
- LSASS access requires context; it is not automatically credential theft.
- Identify the process associated with the LSASS interaction.
- Reconstruct process ancestry before drawing conclusions.
- Validate endpoint-security and administrative explanations.
- Correlate the LSASS event with the existing attack timeline.
- Use process prevalence and signature context as supporting evidence.
- Do not claim credential extraction without evidence.
- Follow the affected account into DeviceLogonEvents.
- Expand scope if the account appears on another device.
- Treat credential-access evidence as a possible bridge from endpoint compromise to lateral movement.
Module 4 — Endpoint Incidents: Following the Attack Chain
Lesson 37 moves the investigation into possible credential access. Next, Agent Foskett follows the affected account when it logs on to another device only minutes later.
Continue your SOC Analyst training
🔎 SOC Analyst Academy — Module 4: Endpoint Incidents: Following the Attack Chain
How to investigate unexpected LSASS access in Microsoft Defender XDR
Lesson 37 of the Agent Foskett SOC Analyst Academy teaches analysts how to investigate unexpected LSASS access, correlate process ancestry and endpoint events, validate legitimate tooling and assess possible credential-access behaviour.
KQL credential access investigation for SOC analysts
Learn how to use DeviceProcessEvents, DeviceEvents and DeviceLogonEvents to connect suspicious endpoint activity with possible credential access and follow-on lateral movement.
