Beginning with the Mark XX, Bolos were equipped with a psychotronic brain which gave them artificial intelligence. Analogously to human minds, psychotronic brains could turn insane when damaged. This could occur during battle, from enemy fire, or from neglect and lack of physical maintenance. It did not occur frequently but, because Bolos were very powerful, the consequences were terrible. For this reason, early psychotronic brains (models XX through XXIII), restricted awareness and initiative at all times except during battle. This was accomplished by separating main processing from personality. The two were integrated, and the Bolo came into possession of its full faculties, only when battle preconditions were met, such as the approach of an enemy or the order of a human officer. In later models, added redundancy reduced the likelihood of insanity and the restriction was relaxed to enhance intelligence. As a final safeguard, Bolos were equipped with a Total Systems Override Program (nicknamed "Omega Worm") which would erase the Bolo's software, rendering it brain-dead. This was triggered if a Bolo refused an authorized order, or it could be executed by a human operator.
The cognitive inhibitions were completely removed after a review of the combat performance, at the Battle of Santa Cruz (c.3030 A.D.), of experimental unit 23/B-0075-NKE (Nike). Nike's performance demonstrated the capabilities and reliability of fully autonomous psychotronics. Nike herself died by Omega Worm as a result of refusing to obey an officer who was a traitor. This led to a revision of the parameters that would lead to the execution of the Omega Worm in later model Bolos.
Beginning with the Mark XXV models, Bolos became completely autonomous, capable of full self direction in all situations. However, it was found that the intuitive capabilities of human commanders working in conjunction with intelligent Bolos increased the effectiveness of the units and so, with some exceptions, human commanders continued to be assigned to, fight with and if necessary, die with their Bolos. This partnership was further enhanced with the introduction of the Mark XXXII which pioneered a neural interface which allowed the Bolo and its human commander to mentally merge human intuition and Bolo processing speed.
Early Bolo models were not self-aware artificial intelligences and up through the Mark IX were only systems which automated the functioning of the vehicle under direct human command. Beginning with the Mark X, Bolos began to use limited AI systems using pre-packaged battle plans which allowed them to function relatively independently provided the situation on the battlefield fell within the parameters of its pre-loaded plan. If not, then the human commander needed to directly intervene either selecting a new battle plan or taking over the functions of the Bolo personally. This system was further advanced beginning with the Mark XV-R which was given a basic AI core capable of choosing between various pre-loaded plans based upon actual battlefield conditions. However, what these earlier Bolos were not capable of doing was developing their own independent battle plans.
Most later mark Bolos have several processing centers, the main core, the personality center, the damage control core (in later models) and the units survival center. Some Bolos were given a secondary main core however this was not typical. Of these the survival center was the most heavily protected being at the center of the Bolo's mass directly under the human commander's command deck. In the event of the destruction of the Bolo the survival center was designed to protect the Bolo's core personality and programming for later retrieval and reactivation.