Massive Malvertising Ambush: One Million Devices Compromised Globally
Summary
- Microsoft says a malvertising campaign reached nearly one million devices worldwide.
- Microsoft Threat Intelligence detected the campaign in early December 2024 and published details on March 6, 2025.
- The attack chain started on illegal streaming sites and used GitHub, Discord and Dropbox to host malware.
- The payloads stole information, including browser data and credentials.
What Microsoft found
Microsoft says a large malvertising campaign hit nearly one million devices around the world. Homes and businesses alike. Microsoft Threat Intelligence detected it in early December 2024 and laid out the full attack chain in a March 6, 2025 report.
How malvertising works
Malvertising combines “malicious” and “advertising.” Criminals use online ads to steer people to harmful sites or downloads. In this case, the ads ran on illegal video streaming sites. Clicking them sent victims through a series of redirect sites before landing them on malware.
Inside the attack chain
Microsoft traced the redirects to malicious files hosted mostly on GitHub, with some payloads on Discord and Dropbox. Familiar platforms helped the attackers blend in. GitHub took the repositories down.
Start your morning with the signal that matters.
Get the weekday cyber brief for the AI risk, regulation, abuse, and enterprise security changes this article could not cover.
By subscribing you agree to our Privacy Policy.
Free. Weekday mornings. 5 minutes or less.
Several stages of malware
The campaign worked in stages. The first payload gathered details about the infected system and sent them back. Later stages installed information stealers, including Lumma Stealer and an updated version of Doenerium, and on some machines the NetSupport remote access tool, while PowerShell, AutoIt scripts and registry changes kept the malware running after a reboot.
Who was affected
The campaign was indiscriminate. It hit individual users as well as organizations across many industries, and stolen browser data and credentials can give attackers a way into corporate accounts long after the original infection has been cleaned up. An earlier version of this article included quotes attributed to a “Microsoft representative” that could not be found in Microsoft’s report; they have been removed.
How to protect yourself
Microsoft recommends hardening endpoints and browsers. Keep software updated, use reputable endpoint protection with cloud-delivered protection turned on, and turn on multi-factor authentication. For companies, blocking access to illegal streaming sites and restricting script engines like PowerShell for ordinary users cuts off much of this kind of attack.
Lessons from the campaign
One click on a pirated streaming site was enough to start the chain, and the attackers hid behind trusted platforms the whole way.
Start your morning with the signal that matters.
Weekday mornings. Five minutes. What changed in security, why it matters.
By subscribing you agree to our Privacy Policy.
Free. Weekday mornings. 5 minutes or less.
That combination is common now. Defenders need to watch for unusual downloads from code-hosting and file-sharing sites. Treat ads on shady sites as a hazard.
Newer CyberExperts coverage on this topic
This article still works as background. If you want the current picture, start with the freshest related coverage below and today's brief.
Citrix NetScaler CVE-2026-19490: The Auth Bypass That Went From PoC to Probes in a Day
CVE-2026-19490 is a CVSS 9.3 NetScaler Gateway/AAA auth bypass. A public PoC on Sept 2 was followed by live probes within a...
PaperCut NG/MF: The Print Server That Learned Remote Code Execution Again
CVE-2026-81578 and CVE-2026-82078 chain to pre-auth RCE on PaperCut Application Server. Confirm Emergency Patch Release 2, restrict admin exposure, and hunt like...
JFrog Artifactory CVE-2026-82329: Empty Join Key, Full Admin Token, Busy Scanners
CVE-2026-82329 lets attackers forge Artifactory admin tokens via a deterministic empty join key. Mass scanning peaked near 406,000 attempts; patch fixed builds...
Friday’s brief: Apache’s HTTP/2 fix, then Check Point break-ins, a ScreenConnect add-on and a GitHub secret stealer
The fastest way to catch up on what changed after this article was published.
Start your morning with the signal that matters.
Get the biggest cybersecurity developments, why they matter, and where to go deeper on CyberExperts.
By subscribing you agree to our Privacy Policy.
Free. Weekdays. Built for operators.