1. Home
  2. Blog
  3. Security, Fraud & Insider Threats

Can Employee Monitoring Detect Unauthorized Logins? What It Sees, Alert Rules and a Response Playbook

Published: Read time: 7 minsAuthor: Subhash Miriyala

Can Employee Monitoring Detect Unauthorized Logins? What It Sees, Alert Rules and a Response Playbook

Summary

Yes, employee monitoring can help detect unauthorized logins, but it works best as one layer alongside your identity provider. Endpoint monitoring sees what happens after someone is signed in to a monitored work device: the time the session started compared with that person's normal pattern, where the device appears to be, which apps and files are touched, and, with behavioral biometrics, whether the typing and mouse rhythm still matches the account owner. Your identity provider, such as Microsoft Entra ID or Okta, sees the sign-in attempts, including failed passwords, MFA prompts and impossible travel. Combining the two catches more: the wrong person at the keyboard, logins at the wrong time or place, shared credentials and accounts that should have been closed. The key is a short list of well-tuned alerts and a clear, fair response process.

E-Monitor's real-time alerts combine rule-based triggers such as geographic mismatch with per-user anomaly detection that flags unusual login times, so suspicious sessions reach the right reviewer quickly. Start free →

What Counts as an Unauthorized Login

An unauthorized login is any session on a company account or device that the organization has not approved, whether or not the password was correct. That definition matters, because many unauthorized logins use valid credentials. The common types:

  • Wrong person: a stolen or phished password, a colleague using someone else's account, or a family member on a work laptop.
  • Wrong time: sessions outside a person's schedule with no on-call or overtime reason, especially on sensitive systems.
  • Wrong place or device: a sign-in from a country the employee has never worked from, or from an unmanaged personal device when policy requires a company one.
  • Shared credentials: two or more people knowingly using one login, which breaks attribution even when nobody intends harm.
  • Dormant or former accounts: an ex-employee, a finished contractor or a test account that was never disabled and is suddenly active again.

Each type leaves different traces in different systems, which is why no single tool catches all of them. Consider a generic example: a 40-person finance team where one analyst's password was phished. The attacker signs in successfully, passes an MFA prompt the analyst approved by mistake, and starts a session at an hour the analyst never works. The identity provider records a successful sign-in from an unfamiliar location. Endpoint monitoring records a session that starts far outside the analyst's usual pattern, opens apps the analyst rarely uses and, if behavioral biometrics is enabled, types and moves the mouse nothing like the analyst. Any one of those signals might be explained away. Together, they make a strong case for immediate review.

What Monitoring Software Sees vs What the Identity Provider Sees

The most useful thing to understand about login security is where each signal lives. Endpoint monitoring and identity logs overlap less than people assume.

SignalIdentity provider or SIEMEndpoint employee monitoring
Failed password attempts and lockoutsYes, primary sourceNo
MFA prompts, denials and fatigue attacksYesNo
Impossible travel between sign-insYes, often a built-in risk detectionPartial, through geographic mismatch on the device
Sign-ins to cloud apps from unmanaged browsersYesNo, unless the device is monitored
Session start time vs the person's normal patternPartialYes, with a per-user baseline
What the user did after signing inLimited to app-level audit eventsYes: apps, websites, files, USB, idle and active time
Whether the person at the keyboard is the account ownerNoYes, with behavioral biometrics

So failed logins, password spraying and impossible travel usually come from identity logs. Microsoft documents the contents of its Entra ID sign-in logs, and Okta's System Log serves the same role. Endpoint monitoring adds the behavioral layer: what the session looked like and whether it matched the real user. Our comparison of employee monitoring vs SIEM explains how the two data sets fit together.

How to Detect Unauthorized Logins with Monitoring Data

To detect unauthorized logins reliably, combine three kinds of signal rather than relying on any single one.

1. Baseline anomalies

Login anomaly detection compares each session with that user's own history instead of a company-wide rule. A night-shift engineer starting at 11 p.m. is normal; an accountant who has never worked past 6 p.m. starting a session at 2 a.m. is not. Per-user baselines cut false positives sharply compared with fixed after-hours rules. Our explainer on anomaly detection in employee monitoring covers how baselines are built and tuned.

2. Context rules

Some conditions are worth an alert regardless of history: a session from a region where the company has no staff, USB storage connected straight after an unusual login, or a dormant account suddenly becoming active.

3. Identity of the person at the keyboard

A valid password and a passed MFA prompt prove that someone had the credentials, not who is typing an hour later. Behavioral biometrics measures typing and mouse timing patterns, never content, and flags sessions that stop matching the enrolled user. It is the main way to catch session hijacking and the quieter problem of colleagues sharing a login, which we cover in how to detect account sharing among employees.

How E-Monitor Flags Suspicious Login Activity

E-Monitor contributes the endpoint side of the picture for monitored work devices:

  • Unusual login time: anomaly detection learns each user's baseline over a rolling 30-day window and fires when behavior deviates, including an unusual login time or uncharacteristic app access.
  • Geographic mismatch alerts: rules can trigger when a session's location does not match what is expected for the user.
  • Attendance records: attendance tracking stamps every clock-in with device, user and time against the employee's shift policy, so a session that starts on an unfamiliar device or at an odd hour is visible in the record.
  • Behavioral biometrics (Professional and Enterprise, opt-in per role): flags sessions that no longer look like the account owner and can request step-up re-authentication through a one-time code, an SSO re-challenge or manager approval.
  • Tamper-evident logs: activity logs are append-only with SHA-256 hash chaining, and export to CSV, JSON or your SIEM so analysts can line them up against identity events.

Two honest limits. E-Monitor does not see failed password attempts or MFA events at your identity provider; those stay in Entra ID, Okta or your SIEM. And because monitoring runs only during clocked-in work hours, it is designed to watch work sessions, not to surveil devices around the clock. For privileged accounts, where the stakes of a bad login are highest, see our guide to privileged user monitoring.

Alert Rule Examples for Unusual Login Detection

Start with a handful of rules like these and tune them against two to four weeks of real data before routing them to on-call staff.

RuleSignal sourceSuggested severityFirst response
Session start far outside the user's 30-day baselineEndpoint anomaly detectionMediumCheck schedule, on-call rota or overtime approval
After-hours login on a finance, HR or admin accountEndpoint plus identity providerHighContact the user through a separate channel
Session location in a region with no staffGeographic mismatch ruleHighConfirm travel; if unconfirmed, reset credentials
Impossible travel between two sign-insIdentity provider risk detectionHighRevoke sessions and require MFA re-registration
Behavioral profile mismatch mid-sessionBehavioral biometricsHighStep-up re-authentication, then review
USB storage connected shortly after an anomalous loginCombined ruleCriticalEscalate to security immediately
Activity on an account marked as departedHRIS status vs activity logCriticalDisable account and preserve logs

The endpoint rules map to E-Monitor's real-time alerts, which support single signals and combinations. The impossible-travel rule belongs in your identity provider, and the departed-account check is usually built in your SIEM by joining HR status with activity and sign-in logs.

A Response Playbook for Suspicious Logins

  1. Triage the alert. Check the obvious explanations first: travel, a schedule change, an approved overtime shift, an outage.
  2. Verify with the user out of band. Call or message them on a channel the session cannot control. Never ask "was this you?" through the possibly compromised account.
  3. Contain if unconfirmed. Revoke active sessions, reset the password, require MFA re-registration and, if needed, isolate the device.
  4. Preserve evidence. Export the activity log and identity events for the window before and after the login. Hash-chained logs show the record has not been altered.
  5. Scope the impact. Review what the session touched: files, apps, USB, cloud storage and data exports.
  6. Close the gap. Fix the root cause, whether a phished password, a shared login, a missing MFA policy or an account that should have been disabled.
  7. Document and tune. Record the outcome and adjust the rule if it fired for a legitimate reason.

Ex-Employee and Contractor Accounts

Dormant accounts are among the most common sources of unauthorized access, and among the easiest to prevent. When someone leaves, disable their identity provider account the same day, remove them from shared mailboxes and admin groups, rotate any shared secrets they knew, and remove the monitoring agent assignment so their records move into the retention policy. Any activity on an account after its owner's end date should be treated as a critical alert. Our guide to offboarding and data retention covers the monitoring side of this checklist.

Contractors and vendors need the same discipline, with one addition: give every engagement an end date in the identity provider so the account expires automatically if nobody remembers to close it. Review dormant accounts monthly, because an account with no activity for 60 days that suddenly signs in is more suspicious than one in daily use.

Avoiding False Positives and Alert Fatigue

An alert that fires every night gets ignored by the second week. To keep login alerts useful:

  • Use per-user baselines instead of company-wide after-hours rules, and give new employees time to build a baseline before strict rules apply.
  • Feed shift schedules, on-call rotas and approved travel into your rules so expected exceptions do not alert.
  • Reserve high-severity routing for combinations, such as an unusual login followed by a sensitive action, rather than single weak signals.
  • Account for VPNs and cloud proxies that make a user appear in a different city.
  • Review every rule's hit rate monthly and retire rules that never produce a real finding.

Our article on employee monitoring alert fatigue goes deeper on severity tiers and routing. Finally, keep the process fair: tell employees in your monitoring policy that unusual logins are reviewed, treat an alert as a question rather than an accusation, and let people explain before any action is taken. Transparent, disclosed monitoring is the approach behind E-Monitor's employee monitoring software.

Frequently Asked Questions

1. Can employee monitoring detect unauthorized logins?

Partly. Endpoint monitoring can flag sessions at unusual times, from unexpected locations, or where the person at the keyboard does not match the account owner. Failed password attempts, MFA denials and impossible travel come from your identity provider. Using both together gives much better coverage than either alone.

2. Can monitoring software see failed login attempts?

Usually not. Failed sign-ins, lockouts and MFA prompts are recorded by the identity provider, such as Microsoft Entra ID or Okta, and by your SIEM. Employee monitoring software sees what happens on a work device after a successful sign-in, so route identity logs and monitoring logs into the same review process.

3. How do after-hours login alerts work?

After-hours login alerts compare when a session starts against a schedule or a learned baseline. Fixed rules flag anything outside set hours, while anomaly detection flags sessions that are unusual for that specific person. Baseline-based alerts produce fewer false positives, especially for teams with shifts, on-call duties or flexible hours.

4. What is login anomaly detection?

Login anomaly detection learns each user's normal login times, locations and session behavior, then alerts when a new session deviates significantly. E-Monitor builds its baseline over a rolling 30-day window and can flag unusual login times and uncharacteristic app access, which helps surface compromised or misused accounts.

5. Can employee monitoring tell if someone else is using my account?

With behavioral biometrics it can raise a signal. The feature measures typing and mouse timing patterns, not content, and flags sessions that stop matching the enrolled user. E-Monitor can then request step-up re-authentication, such as a one-time code or SSO re-challenge, before a reviewer looks at the alert.

6. What should I do if I get a suspicious login alert?

Check for an innocent explanation, then verify with the user through a separate channel. If the login is not confirmed, revoke sessions, reset credentials and require MFA re-registration. Export the relevant logs, review what the session accessed, fix the root cause, and document the outcome.

7. How do I reduce false positives on suspicious login activity alerts?

Use per-user baselines instead of fixed hours, include shift and on-call schedules in your rules, account for VPN location changes, and escalate only on combinations of signals. Review each rule's hit rate monthly and remove rules that never lead to a real finding.

Catch the wrong person, time or place. E-Monitor's anomaly detection, geographic mismatch alerts and behavioral biometrics add the endpoint layer to your identity logs, with tamper-evident records ready for your SIEM. 7-day free trial, no credit card. Sign up →

Ready to see E-Monitor in action?

E-Monitor deploys in minutes with screenshots, productivity analytics, real-time alerts and compliance-ready reporting. 7-day free trial, no credit card required.