logoalt Hacker News

chunkyslink • today at 1:21 PM • 2 replies • view on HN

How the ad-based spyware actually worked (from the Citizen Lab and Confiant reports):

- It came in through the front door. Discover-Tech ran its own ad-buying platform and was listed as a "certified" vendor by Google and IAB Europe for six years. Nothing was hacked to get in. It bid in ad auctions like any other buyer.

- The ads looked harmless. Plain banners for games, AliExpress deals and an IP-lookup site. Hidden inside each one was a small script that quietly phoned home when the ad loaded.

- Everyone who saw an ad was profiled. The script collected device details, IP addresses and, on many sites, the publisher's own cookies, sometimes including logged-in session tokens. Cookie-syncing with other ad networks built a huge identity graph.

- Only chosen targets got the payload. A gatekeeper server checked each device's ad ID and IP. Most people got nothing back. A selected target got code that handed them off to spyware maker Candiru, and the code was served only once, which made it very hard for researchers to capture.

- The handoff was built to defeat inspection. Encrypted channels, rotating hosting, anti-bot checks and even a fake "Copyright Google" comment to fool human reviewers.

- It worked. At least 45 devices at Al Jazeera were targeted, and spyware was recovered from an infected machine. Its ads also appeared on networks belonging to NATO, the UN, the European Parliament, the ICC and various defence ministries.

- The scale was huge. Billions of bids and ads in nearly 200,000 apps and websites, including Duolingo, Viber, Yahoo Mail, dating apps and period trackers, yet almost no sign of real advertisers or revenue.

Citizen Lab: https://citizenlab.ca/research/how-discover-tech-turned-targ...

Confiant: https://blog.confiant.com/p/adint-weaponized-ad-tech-for-zer...

Citizen Lab also has a shorter executive summary: https://citizenlab.ca/research/exposing-adfection-executive-...


Replies

aenis • today at 2:01 PM

Wait what, so there is no content disarm between companies making ads, and the publishers? They can literally push whatever they want? Wild.

Edit:

the TFA answers: "Advertisers or DSPs may run HTML and JavaScript code when an ad is shown to the user." Its wild! I thought the ads are sanitized, go through a CDR of sorts, and whatever javascript is permitted is authored by ad networks, not the buyers. Sheesh. Thats just RCE.

➕ show 2 replies
gruez • today at 2:09 PM

Did you AI generate this? The wording looks suspicious and sure enough, it's flagged by pangram as AI.

➕ show 1 reply