This applies to all of you, no matter what it is you pirate. Software, video, audio, books, bomb-making instructions, whatever.
You lot really need to be doing a better job at concealing your files from your filehosts, especially when you're working at an open port like OneDrive or Google Drive. I feel like this isn't a conversation we need to keep having. You're good enough at computers to know how to make all your archiving tools work, you should be good enough to understand some very basic things about online privacy and file security.
Obviously, Google Drive, Mediafire, and the Internet Archive are easy for us all to access; both as uploaders and downloaders. As liberationists, we want our datahoards to be as easy for people to find as possible, within reason. However, all of these so-called "cloud storage" services are constantly monitored for compliance with international law. Some hosts, like Google, will even have a look inside your archive files to see what's there and if it sounds like something that might get them sued.
The implication here is obvious: don't name your files in plain language. OneDrive is not going to let you keep a file called "Super Smash Bros Ultimate (All DLCs) (Ryujinx emulator optimised) Repack Part 1.7z" on your account. To be quite honest, it probably won't let you keep "U3VwZXIgU21hc2ggVWx0aW1hdGUgcGx1cyB3aGF0ZXZlciBJIHNhaWQ=.7z" either. Base64 decoding is a pretty easy process, which can be completely automated. You should operate under a few assumptions when you upload your hoards to whatever filehost you plan to use:
1. A filename in plain language will be discovered by the host.
2. The host has some process to automatically decrypt Base64 and other common encryption schemes.
3. Analytical AI will be turned onto a random selection of files in order to look for key words and phrases.
As for the actual contents of the files (ie. the actual code that makes up the liberated file), this may be done on files uploaded by accounts that they suspect of piracy, but it would be too costly to do on every single file uploaded to the service. Some of these hosts get literal hundreds of thousands of files uploaded to them every day. To conduct a search for pirate content after a file has been uploaded would be like searching for a needle in a haystack made of other needles. While it might be more economical to conduct the search during the upload procedure, this would slow the process down such that people may not use the service at all. While Google might do this, other filehosts probably won't, at least not until analytical AI becomes cheaper. Even so, automated searches for potentially infringing content have been going on for nearly 20 years without analytical AI. Ask all the people who ripped their Warner Records CDs and uploaded them to YouTube when they were kids about that.
Basically, what I'm getting at here is, obfuscate your filenames. The nearest thing I've seen to anti-detection measures in piracy so far is someone passing off a complete Nintendo 64 collection on the Internet Archive as "Solarix Business Quarterly 1996-2002". At the same time, none of Fitgirl's Nintendo Switch repacks have valid direct download links anymore because the uploaders all named the files in plain language. They called their repacked version of Super Mario Odyssey "Super Mario Odyssey Repack". Yeah. Great job there, bucko. You forgot your balaclava at home, but at least you remembered your brightly lit sign that reads "PIRATED CONTENT HERE".
"So, if not Base64, then what?" You've got to think simpler. Pretend like you're a World War II era spy, trying to transmit information on German troop movements to the British. If you're trying to put a collection of all the N64 Mario games onto the Internet Archive, call it something like "Mushroom Foragers Monthly - 1964". Maybe don't even bother with trying to be clever and just set your naming conventions to "Keysmash". Call it "v ehqnrioe". Call it anything except "Nintendo 64 - All Mario Games".