How MSPs Can Detect Phishing Attacks That Email Filters Miss

· Original article ↗

Summary

A sponsored Kaseya article advises MSPs to look beyond email filters by monitoring account behavior, correlating identity, email and endpoint signals, and quickly containing suspected compromises.

Key points

  • The article says AI can help attackers personalize phishing messages and vary them to evade traditional email filters.
  • Watch for suspicious mailbox rules, logins from unfamiliar locations, impossible travel, and repeated unsolicited MFA prompts.
  • Correlate identity, email and endpoint activity; suspicious processes after login, unusual access attempts, or a sudden rise in outbound email can indicate an active attack.
  • Recommended responses include investigating suspicious account activity, isolating compromised endpoints, disabling affected accounts, and ending active sessions.
  • The article recommends current phishing simulations, separate-channel verification of high-risk requests, post-delivery account monitoring, and measuring detection and containment times.

Article Details

Defense Focus
Detect and contain phishing-related account compromise by correlating identity, email, and endpoint activity beyond the initial email delivery.
Detection Methods
  • Flag new mailbox forwarding rules, especially when followed by a login from an unfamiliar location.
  • Detect impossible travel and repeated multifactor-authentication prompts the user did not initiate.
  • Correlate identity, email, and endpoint alerts, including PowerShell or other unusual processes launched after a login.
  • Identify successful logins followed by unusual access to systems, applications, or data.
  • Monitor for sudden increases in outbound email from accounts that usually send few messages.
  • Use behavioral analytics, anomaly detection, and automated threat correlation to surface related signals.
Data Sources
  • Authentication and identity activity, including login location and device
  • Email and mailbox activity, including forwarding rules and outbound message volume
  • Endpoint process activity and alerts
  • Multifactor-authentication prompt activity
  • Access to systems, applications, and data
Platforms
  • Email gateways
  • Endpoint detection and response
  • Identity monitoring
  • Automated threat correlation
Defensive Actions
  • Run phishing simulations that reflect current AI-generated phishing rather than outdated, generic examples.
  • Verify wire transfers, credential resets, and vendor payment changes by phone or another separate channel.
  • Continue monitoring account activity after email delivery for suspicious mailbox rules, unfamiliar-location logins, impossible travel, and unexpected MFA prompts.
  • Automatically flag and investigate suspicious account activity.
  • Isolate compromised endpoints to limit malware spread.
  • Disable compromised accounts or terminate active sessions to prevent further access.
  • Measure time to detect and contain suspected compromises.

MITRE ATT&CK