If AI is trying to stop spam, wouldn’t AI also be making better, more believable, spam?

Of course it also depends on where AI resources were spent. AI is still fairly new. Has Apple applied it to their spam identification yet? Has your host?