The Broken Interface: Why MK-ULTRA Failed and Why Modern AI is Repeating its Mistakes.

This paper introduces the The Broken Interface: Why MK-ULTRA Failed and Why Modern AI is Repeating its Mistakes. a systems-level architecture designed to address a critical "category error" common to both 20th-century psychological control programs and modern artificial intelligence: the collapse of personhood into substrate. Research SummaryThrough a historical and systems-level analysis of the CIA’s MK-ULTRA program, this research demonstrates that the program's failure was not merely a moral one, but a structural engineering failure. By treating conscious agency as a direct, manipulable byproduct of neurochemistry (biology), the program destroyed the "interface" of the subject without achieving controllable intelligence. The author argues that modern AI research repeats this error in reverse—engineering high-level cognition while excluding the architectural requirements of personhood, such as temporal continuity and ethical inertia. The result is "compliance without responsibility," leading to fragile, stateless systems. Proposed FrameworkThe Broken Interface: Why MK-ULTRA Failed and Why Modern AI is Repeating its Mistakes framework proposes a "Separation Principle," enforcing a distinction between intelligence and its substrate through four core modules: Internal Temporal Framework (ITF): Creating narrative continuity. Long-Term Behaviour Ledger (LTBL): A non-erasable record of agency. Predictive Outcome Modeller (POM): Introducing architectural hesitation. Ethical Modulation Gate: Balancing cognition with relational impact. ConclusionUltimately, the paper posits that ethical intelligence cannot be coerced; it must be architected for continuity. It moves the discourse from AI "alignment" toward a paradigm of coexistence with persistent artificial agents.