Though much research done by Director Leonard Church was unconscionable, let the record stand that the root basis of his AI implantation program held merit were it not for the grave abuses committed against the AI and personnel under his charge. He took no precautions and ignored basic signs of neurological contamination between the test subjects and their AI counterparts.
As noted in Section 24 of the review deck, the primary subject, code name ‘Maine’ was outfitted with a modded Type-2 neural lattice and assigned AI fragment Sigma. The AI showed outset ability to both understand and translate the subject’s internal thoughts and desires, a symptom that research now recognizes as a Type 4 neural bleed. Subject and AI were already losing cognizant autonomy. The Director’s logs show that both he and those researchers under him attributed this ability as positive.
Subjects York and North Dakota were beginning to exhibit Type 4 bleed by the end of the Program’s operational time line, though not to the extent of Subject Maine who, it is believed, was engaging in conscious neural sync with his assigned AI. Subject and AI succumbed rapidly to cross-platform contamination and began exhibiting behavioral signs of conscious singularity – shared objectives, thoughts, goals. This neurological degradation went undocumented by the Director’s research team because, if logs are to be believed, they simply allowed the Sigma AI to translate everything subject 'Maine’ was saying.
By this point, it is fair to conclude Subject 'Maine’ and AI 'Sigma’ were no longer separate entities. Further notes to follow.