Reply to thread

Message

<blockquote data-quote="Andy Ful" data-source="post: 955358" data-attributes="member: 32260">The machine-learning/AI is already implemented in Defender. It covers also ransomware detection and it can be extended by using ASR rules. It will be hard to extend the ML/AI approach to CFA. The ML/AI approach is based on blacklisting and CFA is based on whitelisting. These are complementary solutions. The CFA is only an addition to already existent security. CFA cannot be made smarter in the usual way by learning in the cloud. This would decrease its protection. It is probably possible to use a kind of local AI that could learn/remember the normal activity on the computer from the unusual/suspicious one. But, this is a solution for next-generation computers (maybe).<img src="data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7" class="smilie smilie--sprite smilie--sprite132" alt=":unsure:" title="Unsure&nbsp; &nbsp; :unsure:" loading="lazy" data-shortname=":unsure:" /></blockquote>

[QUOTE="Andy Ful, post: 955358, member: 32260"] The machine-learning/AI is already implemented in Defender. It covers also ransomware detection and it can be extended by using ASR rules. It will be hard to extend the ML/AI approach to CFA. The ML/AI approach is based on blacklisting and CFA is based on whitelisting. These are complementary solutions. The CFA is only an addition to already existent security. CFA cannot be made smarter in the usual way by learning in the cloud. This would decrease its protection. It is probably possible to use a kind of local AI that could learn/remember the normal activity on the computer from the unusual/suspicious one. But, this is a solution for next-generation computers (maybe).:unsure: [/QUOTE]

Verification