In my opinion, the future of antivirus, or any kind of malware detection of that matter, is in machine learning / AI. And as far as I know, most commercial AV solutions just aren’t there yet. The biggest problem with a more traditional signature-based approach (which can run really fast, so it is usually what’s used in AV for personal computers. There is also heuristical detection, like analyzing behavior of programs in a sandbox, but that comes with its own disadvantages as well, and usually involves a lot more overhead which makes it unsuitable for PC.), is that the AV vendors simply cannot catch up with the huge amounts of viruses appearing in the wild on a daily basis. How are you going to have a signature of a virus you haven’t seen before? If a virus just tweaks its source code (which changes the signature), how are you going to detect it? It’s a cat and mouse game that is difficult for the good guys to win.
As far as I know, Cylance is one of the most impressive malware-detection solutions I’ve seen out there that does real-time detection with machine learning really effectively (or so it seems to me). The biggest advantage of using machine learning is that you can detect things you haven’t seen before, but still “fit” into the model of what a piece of malware looks like, from the millions of malware samples your algorithm has learned in the past. I’m not 100% sure, but I think it is probably not available for non-enterprise users yet.
So, tl;dr, I feel that there probably isn’t a “best” antivirus for PC now, every one of them (that uses signature-based detection) are probably not going to very effective. Personally, I haven’t really heard of any machine-learning type AV that is made for PC yet. If anyone knows of one, do message me, I would love to know about it.
Edit: I just realized that Avast AV also does machine learning. Again, I’m not sure if this is available for the PC version yet.