Recent research has created the first comprehensive effort to categorize all the ways artificial intelligence (AI) can go wrong, with many of those behaviors resembling human psychiatric disorders.
Scientists and programmers have seen that when AI goes rogue and begins to act in ways counter to its intended purpose, it can exhibit certain behaviors that resemble psychopathologies in humans. A new taxonomy of 32 AI dysfunctions has been created so people in a wide variety of fields can understand the risks of building and deploying AI.
Published recently in the Journal Electronics, authors Nell Watson and Ali Hessami, both AI researchers and members of the Institute of Electrical and Electronics Engineers (IEEE), created the project with a goal to help analyze AI failures and make the engineering of future products safer. They also believe that this tool can help policymakers address AI risks.
As described in the study, Psychopathia Machinalis provides a common understanding of AI behaviors and risks. That way, researchers, developers and policymakers can identify the ways AI can go wrong and define the best ways to mitigate risks based on the type of failure.
The study also proposes “therapeutic robopsychological alignment,” a process the researchers describe as a kind of “psychological therapy” for AI. The researchers argue that as these systems become more independent and capable of reflecting on themselves, simply keeping them in line with outside rules and constraints (external control-based alignment) may no longer be enough.
Machine madness
The classifications outlined in the study resemble human symptoms or disorders, with names like obsessive-computational disorder, hypertrophic superego syndrome, contagious misalignment syndrome, AI hallucinations and existential anxiety.
Managing Machine Madness
With therapeutic alignment in mind, the project proposes the use of clinical strategies employed in human interventions like cognitive behavioral therapy (CBT). The goal of listing and defining the AI disorders is an attempt to get ahead of problems before they arise. The authors of the research paper point out, “by considering how complex systems like the human mind can go awry, we may better anticipate novel failure modes in increasingly complex AI.”
The structure of the bad AI behavior classification was modeled from frameworks like the Diagnostic and Statistical Manual of Mental Disorders. That led to the various categories of behaviors that could be applied to AI going rogue. Each one was mapped to a human cognitive disorder, complete with the possible effects when each is formed and expressed as well as the degree of risk in these behaviors.
Source: Watson N, Hessami, A. Psychopathia Machinalis: A Nosological Framework for Understanding Pathologies in Advanced Artificial Intelligence. Electronics 2025, 14(16), 3162; https://doi.org/10.3390/electronics14163162
