"Theft" may be wrong, but "abuse" certainly is not. Human interactions in general, and the web in particular, are built on certain set of conventions and common behaviors. One of them is that most sites are for consuming information at human paces and volumes, not downloading their content wholesale. There are specialized sites that are fine with that, but they say it upfront. Average, especially hobbyist site, is not that. People who do not abide by it are certainly abusing it.
> Likewise if enforcing a rule of no scraping is of utmost importance you need to require an API token or some other form of authentication before you serve the pages.
Yes, and if the rule of not dumping a ton of manure on your driveway is so important to you, you should live in a gated community and hire round-the-clock security. Some people do, but living in a society where the only way to not wake up with a ton of manure in your driveway is to spend excessive resources on security is not the world that I would prefer to live in. And I don't see why people would spend time to prove this is the only possible and normal world - it's certainly not the case, we can do better.
Theft is correct but for a different reason.
The #1 reason for all AI scrapers is to replace the content they are scraping. This means no "fair use" defense to the copyright infringement they inevitably commit.