That is an example of phishing, where somebody wants you to try to log in on a fake web page. Even experts will sometimes absent-mindedly do this, so I advise always using the strongest possible two-factor authentication everywhere. On websites that support FIDO aka U2F, using such a device is highly advisable.
So far as I can tell, only Google has figured out how to kill almost all spam and phishing without any noticeable false hits. And they do it with a neural network running on a large cluster of machines. I have a number of email accounts here and there, and all of them either kill almost all spam but also trap legitimate email, or they let quite a bit of spam through. Even Microsoft with all its resources doesn’t get anywhere near Google’s accuracy.
In the Classic Linux environment, we reject about 80–85% of all incoming mail during SMTP, so that it is never accepted, and then the 15–20% that is permitted in then gets further filtered by SpamAsassin. Factoring all this in, rate of uncaught spam and phishing is probably 0.5–1.0%.
You can get the benefit of Google’s technology by using Gmail’s ability to retrieve mail via POP. (Provided you don’t mind the possible privacy implications of using Google, which I know many people do.)
I’m working on better solutions. The hard part is not killing spam and phishing per se, but rather, doing it without false positives.