The last time I used TOR it was almost impossible to do anything on the internet. Every Google search was met with "We detected you are a bot" and every website interaction was blocked by never-ending CAPTCHAs.
It's definitely not the only use case for such services, but if a service provider sees that 90% of traffic shaped a certain way is malicious traffic, it's understandable they will take steps to mitigate that traffic.
ETA: I'm not happy about it because I believe in the value of anonymity, but it is what it is. Here's a Cloudflare blog post talking about the challenges handling Tor traffic, which to their estimate is (a) 94% malicious "per se," so any tooling you do that tries to estimate intent based on origin IP address is gummed up by the malicious signal emanating from the same Tor exit node as your legit traffic and (b) anonymized by design, therefore any attempts you might make to build a reptutation signal for a given client are intended to be thwarted. The result is that a Tor user's traffic looks reputationless to a service like Cloudflare, and you can't just assume reputationless signal is benign (so, CAPTACHAs and "bot-like behavior suspected" walls).