Academy
TermThreat intelligence

Dark web monitoring

Regularly searching underground forums, log markets and leak channels for an organisation’s credentials, data and mentions.

Flow

  1. 1Sources are collected
  2. 2The organisation is matched
  3. 3The finding is delivered
Related families
Lumma Stealer, RedLine Stealer
MITRE ATT&CK
T1597 · Search Closed Sources
Sources
3 sources
01

Overview

Credentials stolen by infostealers change hands outside the organisation, in sales channels. The organisation cannot see this on its own network; it sees it only by searching those channels. Google Cloud (Mandiant)

Law enforcement operations rely on the same sources: in the operations against Lumma and RedLine, the seized infrastructure revealed which accounts had been stolen. Dutch National Police

02

How it works

  1. 1

    Sources are collected

    Log markets, forums and messaging channels are monitored.

  2. 2

    The organisation is matched

    Domains, email domains and sign-in URLs are searched.

  3. 3

    The finding is delivered

    The matching record reaches the security team with the related account and device details.

03

Real-world examples

Snowflake: passwords stolen years earlier

According to Mandiant, a significant share of the credentials UNC5537 used came from past infostealer infections; some dated back to 2020 and had never been changed. Google Cloud (Mandiant)

04

How to detect it

  • New records for the corporate domain, especially remote access, email and cloud sign-in URLs. Google Cloud (Mandiant)
05

What to do

  • Change the password of the account found, end its open sessions and examine the device separately. Google Cloud (Mandiant)
06

MITRE ATT&CK

07

Related families and terms

08

Sources

  1. 1Search Closed Sources (T1597) · MITRE ATT&CK
  2. 2UNC5537 targets Snowflake customer instances · Google Cloud (Mandiant), 2024
  3. 3Operation Magnus · Dutch National Police, 2024-10-28
From the Essential plan

Infostealer Detection

See your firm's attack surface on one screen: how many findings, at which severity, on which domain.