There's lying. There's lying. And there's lying.
I say that because misrepresentation (mis-re-present-ation) of the facts is always a result of the context in which it is originally done. Children may learn to lie at a very early age because telling the truth is not rewarded with kindness; it is persecuted and punished.
Moreover, it is often done under threat of, or during, physical abuse or other harm... including complex combinations of invalidating the child's reality, or being discounted, disclaimed, rejected, invalidated, confused, betrayed, insulted, criticized, judged, blamed, embarrassed, humiliated, ridiculed, denigrated, derogated, victimized, demonized, persecuted, picked on, dumped on, bullied, gaslighted, and/or scapegoated.
When that is repeatedly the case, the child's memories of both the abuse and his response to it are conditioned, socialized and normalized into the brain's default mode network. Done enough times, the conditioning is habituated by becoming a neural network; a physical actuality that will remain in existence unless or until the abuse ends, and there is no longer a need to re-run the neural firing pattern.
Sadly, in children who were the victims of such abuse as described, the neural wiring I have described develops so much density via alternative networking over time... that it becomes a relatively permanent "fixture" in the brain. And -- just like an amygdala or a hippocampus (which are in fact parts of such neural networks) -- every bit as much a fixed "part" of the brain as those two subcomponents. And those neural circuits will fire off and produce behavior whenever the amygdala and hippocampus sense that the past is being re-present-ed.
Theoretically, at least, all forms of behavior can be de-conditioned with the techniques of classical and/or operant re-conditioning. This is done by specialists in "behavior modification." But when the original conditioning is repeated in a substantial number of slightly different contexts, the neural circuitry that produces the (later undesired) response becomes so dense with alternative channeling that the de-conditioning process may be almost impossibly complex.
When that is the case, the de-conditioning behavior modification will very often have to be done in conjunction with the use of dopamine-channel-blocking, neuroleptic medications at low to moderate dose levels vs. the much higher dosages used for schizophrenia and other forms of psychosis.
Additionally, the treatment team may need to enlist the patient in exposure-based reduction of the emotional energy associated with the repeated, original trauma (see section 7c in this post on Reddit's CPTSD sub). But even then, the neural circuits may just be too dense and alternative to get a "100% perfect" result.
Finally: Shame, guilt, worry, remorse, regret and morbid reflection all play a big part in any addictive process cycle including compulsive misrepresentation. When that is the case,
. . . 1) the habitual liar's entire cognitive schema has to be addressed, definitely including embedded, conditioned beliefs he has about the sources of threat and other triggers that cause the amygdala, hippocampus and default mode network to fire off, and
. . . 2) those stored emotions and sensations as they are related to the original trauma in most or all of the specific contexts will have to be accessed and "bled off."