@dangerzone@social.freedom.press
The implementation is closed-source, but the way this usually works is simple:
1. Craft a unique URL in a domain under your control, that serves an 1x1 transparent pixel image.
2. Associate this URL with the document you want to track, and monitor it for any HTTP GET requests.
3. Plant this URL as an external image in the document you want to track.
4. Whenever someone opens this document, their IP address, operating system and access time will be recorded.
@dangerzone@social.freedom.press
Anecdotally, the only major PDF reader that loads external content by default is Adobe Acrobat. Still, there are various implementations out there, so check what's the case with your PDF reader.