In an age where printed periodicals were delivered by subscription, the subscriber information was available (and yes, often tracked by local and federal law enforcement), but not the specifics of what articles were read.
Today, with Web-based document delivery and Javascript instrumentation, the specifics of who reads what articles, time on page, sections read, interactions, shares, and more, are available not just to the publishere but advertisers, any entities hacking into or accessing their systems, app developers, and more.
And, yes, law enforcement, whether under warrant, subpoena, or ... other methods.
Someone usually will have archived the article there.
If you feel a bit more ambitious you could make a bot that runs on a vps somewhere and automatically scrapes news articles.
Archive.is runs Tor through a Cloudfront captcha which fails consistently in my experience.
Just tried it now and works for me.
It is an annoying captcha, it had something like five steps to complete, but I've seen worse. I'd rather this captcha than the one that Roblox uses.