Copyright is a complex subject about which there are many misconceptions.
I was reminded of that this week, thanks to an email from an author who discovered that several of their books were included in one of the databases of pirated works used by the AI company Anthropic for AI training. The author wanted to know whether they were eligible to be part of the gigantic $1.5 billion settlement Anthropic has agreed to pay to compensate writers for its misuse of their intellectual property. (You can read more about the lawsuit, and the settlement, here.)
One of the criteria for eligibility, set by the court, is that copyrights to the pirated works must have been properly registered with the US Copyright Office before Anthropic downloaded the databases. And indeed, the author’s books were all registered in a timely manner…but not with the Copyright Office. Instead, the author used a website called Copyrighted.com, which offers a kind of faux registration using timestamps and its own certificates.






