Such is the way Grice introduces his 'implicature'. He did that PRIVATELY, as it were, in one seminar on "Logic and Conversation" at Oxford in 1965.
When he travelled to the "New World" (well, Harvard), he used it again.
By the 1980s he had coined a further 'term of art': 'disimplicature'.
Grice wants to say that to disimplicate is not "not to implicate". No. You have to have an M-intention to disimplicate.
So, if you say,
i. Macbeth saw Banquo.
we KNOW that 'saw' ENTAILS
ii. Banquo was there to be seen.
But one can be sloppy, say, in an examination on Shakespeare, and use 'see' WITHOUT this entailment. When you do that, you disimplicate. When you don't, you don't.