Gmail uses AI to detect and block out spam. AI may use a wide range of subjacent algorithms, taking into account bags of variables, amongst which, besides the information on the email itself, it may include prior posts and additional data related to you.

Your case is a false positive, and it’s very possible that, the exact same message sent from another person goes through without any hassle. You can probably throw the algorithm off-track by changing bitcoin for something else, or adding additional text to the message itself. The spam AI may not necessarily be targeting the crypto words per se, but in a given context derived from the message’s content/context and profiling it performs.

Fortunately, these false positives do not seem to be cropping-up at a meaningful rate (I have seen just a few similar cases to yours, reported for example by a user using Hotmail, but it’s not a heavily reported issue). It’s a bummer though when it occurs.

See Google spam filter info: https://www.pepipost.com/blog/gmail-spam-filters-evolution/