Tactical deception

Instinctive dishonesty and learned dishonesty

Even if instinctive dishonesty is not evolutionary stable, innovative dishonesty might be
Individual senders' dishonest signals might derive from (i) operant conditioning (ii) insight (in the sense of Köhler)
Insightful dishonesty is more likely to be effective - if the sender is slowly learning to be dishonest, the receiver can learn equally quickly to discount the dishonest signal (an "arms race" of deception)

Whiten & Byrne noted examples of apparently insightful dishonesty in savannah baboons
E.g. young baboon whose mother is out of sight sees a neighbour eating something attractive and gives a call as if being attacked. Mother attacks neighbour who flees, dropping food; youngster picks up food (so the mother is used as a "social tool")
Whiten & Byrne collected anecdotes of such instances, and found a considerable number in both monkeys and great apes; commonest form is giving an alarm call when no danger is near, with the effect of distracting another animal from food or a mate
Such deception can be called "Machiavellian intelligence"

As with all other acts, there can be controversy over whether deceptive acts are "intentional" or just the result of "blind conditioning"
However intention comes into deception in another way. Some deceptive acts seem to depend on an understanding of the intentions of a social partner (specifically, of the individual to be deceived). This requires the deceiver to have some kind of Theory of mind...