{"id":34775,"date":"2022-01-24T16:30:57","date_gmt":"2022-01-24T21:30:57","guid":{"rendered":"https:\/\/mjtsai.com\/blog\/?p=34775"},"modified":"2022-01-25T16:33:41","modified_gmt":"2022-01-25T21:33:41","slug":"google-drive-incorrectly-flags-file-for-copyright-infringement","status":"publish","type":"post","link":"https:\/\/mjtsai.com\/blog\/2022\/01\/24\/google-drive-incorrectly-flags-file-for-copyright-infringement\/","title":{"rendered":"Google Drive Incorrectly Flags File for Copyright Infringement"},"content":{"rendered":"

Emily Dolson<\/a> (via Hacker News<\/a>):<\/p>\n

Uh, \n@googledrive, are you doing okay? This file literally contains a single line with the number “1”.<\/p><\/blockquote>\n

Presumably nobody manually added the hash of that data to a list, so this is probably some sort of AI mistake.<\/p>\n

The worst part is that the error message says: “A review cannot be requested for this restriction.”<\/p>\n\n

ChicagoBoy11<\/a>:<\/p>\n

\n

I experienced something similar building an internal tool on GSuite. I had a large file with sequences of 9 digit numbers specific to our use-case, all tied to names of people (employees). Whelp, at one point the tool I was working on stopped working, and it was flagged as apparently containing social security numbers (which I suppose matched the character length).<\/p>\n

Whelp, on the admin panel, you can get a report of those files, and then mark it as a false positive. Which I did. But then nothing happened, and nothing changed. It was no use.<\/p>\n

The hilarious bit: It did, of course, allow me to make a copy of the file in question, and then just point the resource I was building to the new file, which was exactly the same. Weeks later... so far, so good.<\/p>\n<\/blockquote>\n\n

Previously:<\/p>\n