There are plenty of things on the web where you can upload data and they don't use machine learning algorithms to decide if your content is "valid" or "invalid". It all depends on various tradeoffs.
I'm pretty sure this isn't a DMCA takedown case here. Google scanned the content of the file, didn't like it and removed access via an algorithm. Nobody but Google knows how it works.
If the content is encrypted then it's impossible to use google docs document viewers, editors, or collaborate online on a document directly via google doc. It becomes just a dumb file host with no added value.
Which begs the question: Is doing things at Google scale really something to aspire to? Has even Google solved the problem of doing things at Google scale if they have problems they cannot solve adequately at Google scale?