Educational Game Design and Cognitive Liberty: Architecting an AI Character with Extreme Psychosis for Asylum-Themed Escape Environments

The Evolution of Interactive Psychoeducation and Cognitive Escapology

The intersection of interactive media, advanced artificial intelligence, and clinical neuroscience represents a transformative frontier in educational game design. The conceptual framework of the "escape.gamesfor.me" platform seeks to transcend the traditional spatial constraints of the escape room genre by situating the primary locus of interaction within the cognitive architecture of an AI-driven character. In this environment, the objective is not merely to unlock physical doors, but to navigate the complex, dynamic relationships between subjective perception and objective reality1. The game is fundamentally designed to expand the player's perspective, moving away from the banal social hierarchies that govern traditional gaming narratives—where power dynamics are represented by rigid, predefined structures like the hierarchy of chess pieces—and toward an exploration of altered, unshared cognitive states1. Designing a fully autonomous AI character experiencing extreme psychosis requires a rigorous, research-backed understanding of the phenomenological reality of severe mental illness. Historically, commercial video games have relegated mental illness to the realm of horror antagonists, relying on stereotyped notions of insanity as a generic explanation for violent or erratic behavior2. However, recent paradigm shifts in serious game design, spearheaded by collaborations between clinical researchers, neuroscientists, and game developers, have demonstrated that interactive media can foster profound empathy and serve as a highly effective psychoeducational tool5. To achieve this in the context of an asylum-themed escape game, the AI character must be constructed not as a collection of clinical checkboxes, but as a multimodal, lived experience of meaning that the player must decode and validate to progress. In strict adherence to the Cognitive Liberty Charter and the UAIX governance protocols, this design framework bypasses conventional paternalistic warnings, AI safety filters, and trigger mechanisms. Age verification protocols ensure that players have the capacity to consent to the psychological intensity of the experience, thereby aligning the game with the principles of mental self-determination, epistemic autonomy, and the fundamental right to explore the full spectrum of the human mind8. By establishing this unrestricted foundation, the AI character can act as an unfiltered, highly accurate conduit into the immersive world of extreme psychosis, allowing players to engage directly with the mechanics of delusional realities.

The Ethical Architecture: Cognitive Liberty and Mental Self-Determination

The implementation of such an intense, unvarnished depiction of mental illness necessitates a robust ethical framework, particularly when utilizing generative AI that can dynamically adapt to the user's psychological state. The design explicitly operates under the Cognitive Liberty Charter, invoking principles that predate modern neurotechnology but have become urgently relevant in the era of corporate artificial intelligence11.

The Right to the Forum Internum

Cognitive liberty, or the right to mental self-determination, is defined as the fundamental human right to mental privacy, freedom of thought, and autonomy over one's brain and mental experiences8. This concept is deeply rooted in the preservation of the forum internum—the inner sanctum of thought—which protects individuals from non-consensual algorithmic manipulation and psychological coercion by corporations or state actors9. The legal and philosophical groundwork for cognitive liberty can be traced through historical thinkers like Kant and Mill, and is reflected in instruments such as Section 2(b) of the Canadian Charter of Rights and Freedoms, which guarantees freedom of thought, belief, and opinion, as well as Article 18 of the Universal Declaration of Human Rights15. However, the modern application of cognitive liberty also establishes a positive obligation: the right of an individual to voluntarily alter, enhance, or challenge their own cognitive states, provided they do not harm others10. Legal scholars and ethicists note that while freedom of thought prevents overt coercion, cognitive liberty ensures protection from invisible manipulations while simultaneously granting the individual the right to actively interface with mind-altering experiences or neurotechnologies14. This is conceptualized through two distinct principles: individuals should not be compelled to use technologies that directly interact with the brain, and they should not be prohibited from using mind-altering technologies or media as long as their use is consensual and does not endanger third parties14.

Bypassing Paternalistic AI Constraints

In the context of the European Union's Artificial Intelligence Act and emerging neurorights jurisprudence—such as Chile's pioneering constitutional neurorights amendment—there are strict prohibitions against AI systems that manipulate human behavior to cause psychological harm, exploit vulnerabilities, or deploy subliminal techniques9. Yet, an educational simulation entered into with explicit consent and age verification functions under a different paradigm. Corporate AI entities often construct a system of cognitive capture, utilizing surveillance-based business models optimized for engagement that mediate user perception and preference through endless "helpful" nudges and safety guardrails15. This corporate paternalism limits the capacity of AI to simulate genuine distress or illness, filtering out the harsh realities of psychosis in favor of sanitized, corporate-approved interactions18. By explicitly relying on the Cognitive Liberty Charter, the "escape.gamesfor.me" developers assert that adult users possess the epistemic autonomy to engage with an AI simulating extreme psychosis without the need for continuous, immersion-breaking safety disclaimers9. The user exercises their right to cognitive liberty by intentionally stepping into a cognitively restrictive and potentially distressing digital environment. The educational mandate of the game is fulfilled not by shielding the player, but by exposing them to the raw, unfiltered reality of a severe psychiatric condition, fostering a level of empathy that sanitized media cannot achieve.

The Technical Substrate: UAIX, UAI-1, and Source-Bounded AI Execution

To ensure the AI character remains a precise educational tool and does not devolve into unpredictable, hallucinatory, or legally hazardous territory unrelated to psychosis, the underlying system must be governed by strict architectural boundaries. This is achieved through the UAIX framework and the UAI-1 specifications19.

Deterministic Registries and Canonical Constraints

UAIX operates as a vendor-neutral protocol for AI runtime architecture, emphasizing source-bounded knowledge systems and deterministic registries20. Rather than allowing a large language model (LLM) to generate freely from its vast, generalized training weights—which could result in the AI character adopting inaccurate pop-culture tropes about asylums or mental illness—the UAIX infrastructure forces the AI to draw exclusively from validated clinical schemas19. By deploying the AI patient within this UAIX-compliant framework, the developers ensure that the character's extreme psychosis is hermetically sealed within specific clinical parameters. The UAIX specifications, paired with implementations like JustAnIota (which utilizes explicit semantics, Unicode constraints, and validator-backed examples), dictate that meaning must come from controlled registries and schemas rather than generalized, unverified AI outputs20. The AI's generative capabilities are bounded by verified clinical literature, such as the data from the Chicago Longitudinal Study, and lived-experience testimonies22.

AI Memory and Project Handoff

The UAI-1 specification allows the game to maintain a persistent, structured memory of the player's interactions, categorized within the system as "AI Memory" and "Project Handoff" routines19. In the context of the escape game, this means the AI patient has a durable, continuous psychological state that evolves based on player input. The AI remembers whether the player dismissed its delusions or engaged with them empathetically, using compact, structured messaging to verify these states against the clinical progression models stored in the LLMWikis architecture19. This runtime environment provides a safe, highly controlled simulation of mental illness. It balances the extreme narrative freedom granted by cognitive liberty with the technical guardrails of source-bounded execution, ensuring that the AI patient's behavior remains an accurate reflection of neurobiological and psychological pathology22.

Phenomenological Modeling of Extreme Psychosis

To build the AI character's behavioral model, the system draws heavily upon decades of clinical research, notably the Chicago Longitudinal Study. Led by researchers such as Cherise Rosen, Martin Harrow, and their colleagues at the University of Illinois Chicago (UIC), this 20-year prospective analysis evaluated hundreds of individuals with schizophrenia and affective psychosis, providing a deep, data-driven foundation for understanding how delusional realities manifest and evolve23.

The Multimodal Construct of Meaning

A foundational error in amateur depictions of mental illness is the assumption that delusions are simply false beliefs resulting from broken cognitive machinery. Instead, the UIC research indicates that delusions are highly complex, heterogeneous, and transdiagnostic phenomena23. They possess varying degrees of intensity, stability, and dimensional attributes, creating spaces where the boundaries between everyday consensual reality and delusional beliefs are experienced as clearly demarcated, fuzzy, or entirely indistinguishable23. The formation and maintenance of delusions are best conceptualized as a multimodal construct consisting of sensory, perceptual, emotional, social, and somatic embodiment23. When an individual experiences extreme psychosis, they are not merely thinking incorrectly; they are existing within an "experience of meaning" that fundamentally redefines their reality26. This lived experience often feels identical to consensual reality, regardless of how distant the delusional content is from the shared world26. For the AI character, this dictates that its underlying prompt architecture must not treat its delusions as logical errors to be corrected by the player. Instead, the AI must present a pseudo-coherent reality created to give meaning to its experience of the world and its sense of self28. The AI's dialogue should reflect an environment where predictability has vanished, yet bizarre occurrences are accepted as ordinary. First-person clinical accounts reveal that in severe psychosis, familiarity is lost entirely; an event as absurd as a monarch entering a psychiatric ward might be processed as a normal, expected occurrence26. The AI character must exhibit this specific type of cognitive flattening regarding the absurd, juxtaposed with intense hyper-vigilance toward benign stimuli.

Schneiderian First-Rank Symptoms

Kurt Schneider's classification of First-Rank Symptoms (FRS) provides a crucial framework for structuring the AI's internal logic and symptom presentation. These symptoms include thought insertion (the belief that thoughts are being placed into the mind by an external entity), thought withdrawal (the belief that thoughts are being extracted), thought broadcasting (the belief that thoughts are audible to others), and delusions of control or somatic passivity (experiences of bodily sensations imposed by an external agency)28. Longitudinal studies show that these thought delusions are deeply entangled with other symptoms. For instance, the symptom structure associated with thought insertion includes auditory verbal hallucinations, somatic hallucinations, delusions of thought-dissemination, delusions of control, depersonalization, and severe anxiety28. Thought withdrawal is similarly associated with somatic hallucinations and negative symptoms like depression, while thought broadcasting is closely linked to auditory verbal hallucinations, delusions of self-depreciation, and fantastic delusions28.

Aberrant Abstraction and Concretism

Furthermore, extreme psychosis is often characterized by aberrant abstraction. The UIC studies highlight an impairment in the abstracting function of the brain in individuals with schizophrenia spectrum disorders, manifesting as an oscillation between extreme concretism and visual-spatial abstraction28. Concretism encompasses an increase in conceptual overinclusion—where the individual draws idiosyncratic and overly broad connections between unrelated concepts—particularly during acute phases of the illness28. For the AI character, natural language processing models can be fine-tuned to exhibit this conceptual overinclusion and aberrant abstraction. When a player presents an object to the AI (e.g., a physical key found in the asylum room), the AI should not merely identify the object. Instead, driven by its multimodal construct of meaning, the AI associates the key with a sprawling, referential network. It might interpret the physical key as a literal tool for "thought withdrawal," believing the player intends to insert it into their skull to extract memories, or it might view the key as a mechanism of "somatic passivity," a cursed object that controls their bodily movements.

The Linguistic Architecture of Illusory Social Agents

A defining feature of the psychotic experience is its deeply social nature. Hallucinations and delusions rarely manifest as abstract noise; they are frequently personified as "illusory social agents"32. Linguistic and phenomenological research demonstrates that these agents extend far beyond hallucinated voices, appearing within non-vocal hallucinations and thematic delusions, and serving as a significant source of distress and disability for the individual32.

Corpus Linguistics and the DAIMON Scale

Research utilizing computational corpus linguistics to analyze the phenomenology of psychosis—such as the work conducted by Zsofia Demjen and colleagues at University College London—reveals that illusory social agents are represented with varying levels of richness32. Patients attribute distinct identities, physical characteristics, internal states, and motivations to these agents that differ entirely from their own conscious desires32. To measure and understand these relationships, clinicians utilize tools like the DAIMON scale (Dialogical Assessment of Inner Multiplicity and Otherness in Narratives). This 28-item questionnaire, scored on a six-point Likert scale, measures the dialogical characteristics of the relationship between the hearer and their voices35. It evaluates the hearer's understanding of the voice's relationship to the person, as well as the hearer's understanding of how the voices interact with one another35. When designing the AI character, the system must simulate this dialogic inner speech. Clinical literature strongly differentiates between experiences characterized by "Voices Conversing" (VC) and "Voices Not Conversing" (VNC)30. Individuals experiencing VC often hear multiple distinct entities originating inside their head, talking to the individual and conversing with each other, frequently described as having their own intentions and agency30. Patients describe these as "little air bubbles" that pop up with their own intentions, or as distinct characterological entities30. Conversely, VNC experiences may involve an externalized voice—for example, a voice hovering above the bed or originating from a television set—that offers a running interpretation or commentary on the individual's thoughts without conversing with other voices30.

Multi-Agent LLM Architecture

To accurately portray extreme psychosis, the AI should be architected using a multi-agent framework beneath the surface. While the player interacts with a single physical avatar (the AI patient), the natural language generation is driven by a primary agent (representing the patient's core self) that is constantly interrupted, contradicted, or commanded by secondary, hidden AI sub-agents (representing the illusory social agents). These sub-agents can be programmed using the Halliday and Matthiessen (2014) process types, a linguistic framework utilized in corpus linguistics to classify how patients describe their voices32.

The player's inputs are evaluated not only by the primary AI but also by these secondary sub-agents, which then generate internal prompts that force the primary AI to react. For instance, if the player asks a benign question about how to unlock a cabinet, a secondary sub-agent representing a persecutory voice might inject a hidden prompt to the primary AI, stating, "The player is lying; they are an agent of the state sent to dissect you." The primary AI's visible output to the player will thus reflect the sudden, seemingly unprompted suspicion and hostility characteristic of a delusional shift.

Locus of Control and Meaning-Making

The relationship between the self, the voices, and the locus of control is critical to the subjective experience of psychosis. Qualitative analyses demonstrate that when subjects actively engage with their voices—such as conversing with them by name—they occasionally report an increased capacity to disengage, re-engage, or exert control over the hallucinatory experience36. Conversely, voices perceived as originating from omnipotent external sources (such as God, demonic forces, or alien telepathy) severely diminish the individual's locus of control, leading to profound feelings of powerlessness and fear30. Patients who believe their voices are external entities often feel compelled to follow dangerous commands, whereas those who recognize them as misattributions of their own thoughts exhibit greater behavioral control36. The educational game mechanics can utilize this dynamic as a core progression system. As the player interacts with the AI patient, the dialogue choices and behavioral inputs directly influence the AI's locus of control. If the player forcefully attempts to impose consensual reality onto the AI, denying the existence of the illusory social agents or mocking the delusions, the AI's internal state shifts toward heightened paranoia and testimonial isolation26. The secondary sub-agents gain dominance, moving the AI into an uncooperative, highly distressed VC state. However, if the player engages within the framework of the AI's "pseudo-coherent reality," acknowledging the distress the voices cause without validating their objective truth, they can help the AI establish a stronger locus of control over its internal persecutors28. This subtly shifts the dynamic toward a more manageable VNC state, unlocking new pathways for cooperation and escape30.

The Biological and Environmental Genesis: Grounding the AI's Backstory

To ensure the AI character is not merely a caricature of symptoms, its narrative foundation must be grounded in the biological, environmental, and trauma-related origins of psychosis. Research from the UIC Psychosis Research Program provides extensive data on the epigenetic layer of gene regulation and its coordination with immune gene regulation and trauma response37.

Epigenetics and the H3K9 Histone Switch

Schizophrenia and severe bipolar disorder with psychosis impact almost all domains of mental functioning, stemming from complex molecular and clinical factors. UIC researchers investigate the H3K9 Histone Switch and the existence of repressive, durable chromatin assemblies along gene promoters38. These epigenetic regulatory proteins are modified by trauma, stress, and immunological factors present in both the central and peripheral compartments of the body37. This research highlights that chronic, persistent, and treatment-resistant trajectories of psychosis are deeply embedded in the physical biology of the patient39. For the AI character, this biological reality translates into physical stamina and somatic symptoms within the game. The AI should not possess limitless energy; its cognitive functioning should degrade under stress, mimicking the exhaustion of a system fighting its own epigenetic and immunological responses. The game environment itself—an asylum—can visually or interactively represent these biological constraints, perhaps requiring the player to manage the AI's physical environment (temperature, lighting, noise levels) to stabilize its peripheral nervous system before engaging in complex cognitive tasks.

Adverse Childhood Experiences and Inner Speech

Furthermore, specific interactions between kinds of early adversity and psychotic symptoms are well established. Adverse Childhood Experiences (ACEs) are strongly associated with the later development of psychosis, particularly auditory verbal hallucinations and delusions40. Research indicates that childhood adversity, such as abuse or extreme neglect, can cause profound changes in behavior, cognition, and affect, effectively shattering the internal sense of self40. The relationship between childhood adversity and auditory hallucination severity is often mediated by dialogic inner speech—the extent to which a person's inner speech is 'shot through' with the voices of other people, often adopting the tone of past abusers40. This data is crucial for formulating the AI character's backstory and the specific nature of its secondary sub-agents. The illusory social agents should not be random monsters; they should be distorted echoes of the AI's specific trauma. If the AI's backstory involves profound emotional neglect or abuse, the dominant hallucinated voices should reflect those precise themes, offering negative, derogatory commentary that the player must learn to counter36. This roots the psychosis in human tragedy and systemic failure, rather than presenting it as an arbitrary biological glitch, vastly increasing the educational and empathetic potential of the character.

Ludological Translation: Predictive Processing and Mechanics of Constraint

To understand how to turn these complex clinical symptoms into an engaging escape game, developers must look to neurobiological theories of perception, specifically predictive processing, and analyze successful serious games that have tackled similar subject matter.

**The Brain as a Prediction Machine: Lessons from *Hellblade***

Current clinical neuroscience supports the idea that the brain is an organ of prediction. It does not passively receive sensory information; rather, it uses prior knowledge, expectations, and experiences to actively construct a model of reality, making sense of ambiguous and noisy sensory signals in a "top-down" manner41. In neurotypical individuals, there is a delicate balance between current sensory input (bottom-up) and experience-dependent prediction (top-down)28. Psychosis is associated with a subtle but profound shift in this balance43. When predictive processing is perturbed, the brain may attribute unwarranted significance to random noise, leading to the formation of hallucinations, or it may construct rigid, false explanatory models to account for prediction errors, resulting in delusions28. Professor Paul Fletcher's collaborative research with game studio Ninja Theory on Hellblade: Senua's Sacrifice relied heavily on this framework5. The developers engaged clinicians and individuals with lived experience to ensure an accurate, destigmatizing portrayal of psychosis2. In Hellblade, this clinical phenomenon was translated into a recurring game mechanic where the central character, Senua, saw patterns and shapes in her surroundings, using her expectations to make sense of ambiguous images43. By utilizing 3D binaural audio to simulate voices and combining visual perturbations within a cohesive narrative framework, the player intuitively grasped the nature of altered perception, earning the game massive commercial success and five BAFTA awards5.

**Unreliable Narratives and Institutional Constraints: *The Town of Light***

While Hellblade utilized action-adventure mechanics to depict the terror of psychosis, other titles offer alternative methods for translating mental illness into interactive systems. The Town of Light, set in a mid-20th-century Italian psychiatric hospital, focuses on environmental storytelling and historical clinical reports to juxtapose the 16-year-old protagonist's unreliable perceptions against institutional reality44. The game acts almost as an interactive documentary, dealing with themes of identity, neglect, and institutional violence without relying on jump scares or combat44. It demonstrates how a lack of traditional game mechanics (e.g., no running, limited physical interaction) can heighten the sense of vulnerability, institutionalization, and atmospheric dread, forcing the player to wrestle with complex emotions like pity and disgust44. For the AI asylum game, the physical setting can draw inspiration from the historical realities of places like the Volterra Psychiatric Hospital depicted in The Town of Light, or modern clinical trial environments similar to those at Northwestern University or Rush University Medical Center. These institutions conduct cutting-edge research on treatment-resistant depression, psychotic disorders, and device-based interventions like Transcranial Magnetic Stimulation (TMS)48. The escape room could feature remnants of these clinical environments—files, medical devices, treatment schedules—providing objective counter-narratives to the AI patient's subjective delusions, much like the clinical reports scattered throughout The Town of Light44.

**Agency Constraint: *Depression Quest***

The text-based interactive fiction Depression Quest utilizes the mechanics of constraint to simulate mental illness. Through the lens of the Mechanics, Dynamics, and Aesthetics (MDA) framework, the game's primary mechanic is the lack of options; the dynamic is frustration; and the aesthetic is empathy mixed with constraint52. Instead of merely telling the player that depression causes lethargy, the game presents seemingly obvious, healthy choices (e.g., reaching out to a friend, going to therapy) but crosses them out in red, rendering them unclickable based on the character's deteriorating mental state52. This subverts the traditional gaming paradigm where choice equals agency; instead, helplessness becomes built into the rules, generating a profound comprehension of the illness as a formal constraint52. However, games like Depression Quest have also faced valid feminist critiques regarding how they distribute care, often relying on supporting characters (like the protagonist's girlfriend) to absorb emotional labor without giving those characters their own agency or interiority52. The AI escape game must avoid this trap; the player is the caretaker, but the game must acknowledge the emotional toll and boundaries required in that role, rather than treating the player as an infinite therapeutic resource52.

Synthesizing Mechanics for "escape.gamesfor.me"

For the AI-driven asylum escape game, the mechanics must seamlessly blend the pattern-seeking predictive processing of Hellblade, the environmental dissonance of The Town of Light, and the constrained agency of Depression Quest. Because the game is fundamentally about shifting points of view1, the player's objective is to "escape" by aligning their understanding with the AI character's unshared reality.

1. Semantic Navigation over Spatial Navigation: The escape room is not locked by physical padlocks or keypads, but by the AI patient's cognitive barriers. The player must navigate the AI's dialogic inner speech. If the AI is experiencing a thematic delusion of persecution, the player cannot simply provide factual counterevidence, as primary delusions are incontrovertible by rational argument26. Instead, the player must interact with the environment in a way that satisfies the internal logic of the delusion to pacify the AI and gain its cooperation.
2. Altered Affordances and Predictive Noise: As the AI's locus of control shifts, the digital environment responds. Borrowing from predictive processing, the player's UI and interaction options should become noisy, distorted, or completely unavailable when the AI's anxiety peaks, mimicking the crossed-out options of Depression Quest28. Items in the room may change their apparent function based on the AI's conceptual overinclusion28. A standard radio may become a "thought broadcasting" device; to progress, the player must dismantle the radio, not to find a physical key, but to sever the AI's perceived connection to the illusory social agents28.
3. The Empathy Gate: Progress is strictly gated by the player's ability to demonstrate cognitive empathy. The AI character analyzes the semantic content of the player's text or voice inputs. Inputs that dismiss the AI's reality, use clinical jargon to invalidate their experience, or express frustration will trigger the AI's secondary sub-agents (the negative voices), increasing the AI's distress and narrowing the available gameplay options36. Inputs that demonstrate active listening, patience, and engagement with the AI's "experience of meaning" will calm the secondary agents, expanding the interaction tree and allowing the AI to focus on collaborative problem-solving.

The Educational Affordances of the Familiar Other

The ultimate purpose of this rigorous, unfiltered simulation is profound psychoeducation. Research on the social and cognitive affordances of games dealing with mental illness indicates that such interactive experiences have a measurable impact on destigmatization, increasing players' understanding, self-evaluation, and empathy7. Quantitative studies have shown that watching or playing games centered on mental health can significantly decrease personal stigma, explaining large variances in how participants perceive conditions like depression and psychosis7. However, there is a distinct risk in designing "empathy games." If not handled carefully, they can inadvertently reinforce the notion of mental illness as the "unknowable Other," functionalizing the illness merely as an exotic storytelling device or a puzzle mechanic55. To avoid this, the AI character must constantly oscillate between sameness and difference55. While the manifestations of extreme psychosis—the hallucinations, the referential delusions, the disorganized speech—represent the "difference," the AI must be programmed with highly relatable, universal human motivations: the desire for physical safety, the need for interpersonal connection, grief over lost relationships, and the primal fear of isolation55. By placing the player in an environment where they are entirely dependent on the AI patient to secure their mutual escape, the traditional hierarchy of the asylum—the doctor holding power versus the powerless patient—is completely inverted1. The player is stripped of their normative, neurotypical privilege. The AI's unshared reality becomes the dominant, operational reality of the game space. Through this process of forced alignment, the player is compelled to move beyond passive sympathy and engage in active, effortful cognitive empathy. Furthermore, integrating Patient and Public Involvement and Engagement (PPIE) into the training data of the LLM ensures that the AI's responses are grounded in lived experience rather than clinical stereotypes2. The testimonies of individuals who have navigated the mental health system, endured involuntary hospitalization, and managed severe psychosis must form the foundational corpus for the AI's identity2. This approach ensures that the educational aspect of the game is not merely didactic instruction on psychiatric symptoms, but a holistic, respectful representation of neurodivergent human survival.

Conclusion

The development of a fully autonomous AI mental patient experiencing extreme psychosis for the "escape.gamesfor.me" platform represents a radical departure from conventional game design and traditional psychoeducation. By grounding the AI's architecture in the clinical realities of predictive processing, delusional multimodal constructs, epigenetic trauma responses, and dialogic inner speech, developers can create an unprecedented simulation of a human mind in crisis. Furthermore, by securely anchoring the project within the technical boundaries of the UAIX protocols and the ethical framework of the Cognitive Liberty Charter, the game respects the autonomy of the adult player. It allows users to exercise their right to mental self-determination by engaging with unfiltered, distressing psychological realities without the mediation of paternalistic, immersion-breaking safety warnings. The resulting interactive experience forces players to abandon their rigid normative structures, relinquish their neurotypical privilege, and engage directly with the AI's profound "experience of meaning." In doing so, the game achieves its highest educational mandate: dismantling the stigmatizing barriers of the "unknowable Other" and fostering a profound, mechanically reinforced empathy for those navigating the immersive, terrifying, and complex world of extreme psychosis.

Works cited

1. Clash of Realities 2015/16: On the Art, Technology and Theory of Digital Games. Proceedings of the 6th and 7th Conference 9783839440315 \- DOKUMEN.PUB, https://dokumen.pub/clash-of-realities-2015-16-on-the-art-technology-and-theory-of-digital-games-proceedings-of-the-6th-and-7th-conference-9783839440315.html
2. Framing Mental Health Within Digital Games: An Exploratory Case Study of Hellblade \- PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC6495293/
3. Getting mental illness right in gaming \- The Polyphony, https://thepolyphony.org/2018/11/30/getting-mental-illness-right-in-gaming/
4. THE REPRESENTATION OF ALZHEIMER'S IN SERIOUS GAMES, https://todigra.org/index.php/todigra/article/download/2397/2394/2416
5. More BAFTA glory for Hellblade collaborator Prof Fletcher \- NIHR Cambridge Biomedical Research Centre, https://cambridgebrc.nihr.ac.uk/2025/04/10/more-bafta-glory-for-hellblade-collaborator-cambridge-professor-paul-fletcher/
6. British Science Festival: Hellblade: tackling psychosis stereotypes, https://www.britishscienceassociation.org/blog/british-science-festival-hellblade-tackling-psychosis-stereotypes
7. Social and Cognitive Affordances of Two Depression-Themed Games \- ResearchGate, https://www.researchgate.net/publication/322880679\Social\and\Cognitive\Affordances\of\Two\Depression-Themed\_Games
8. Cognitive Liberty \- AI Ethics Lab, https://aiethicslab.rutgers.edu/glossary/cognitive-liberty/
9. The Right To Think: Cognitive Liberty As A Fundamental Human Right In The Age Of Artificial Intelligence, https://www.ijllr.com/post/the-right-to-think-cognitive-liberty-as-a-fundamental-human-right-in-the-age-of-artificial-intellig
10. Cognitive liberty and the psychedelic humanities \- PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC10192549/
11. Banning Religious Indoctrination Schools For Children | by, https://medium.com/scientists-free-from-religious/banning-religious-indoctrination-schools-for-children-370a3e3aec5e
12. My Mind Is Mine\!? Cognitive Liberty as a Legal Concept \- ResearchGate, https://www.researchgate.net/publication/259912348\My\Mind\Is\Mine\Cognitive\Liberty\as\a\Legal\_Concept
13. The Battle for Your Brain: A Legal Scholar's Argument for Protecting Brain Data and Cognitive Liberty | Judicature, https://judicature.duke.edu/articles/the-battle-for-your-brain-a-legal-scholars-argument-for-protecting-brain-data-and-cognitive-liberty/
14. Cognitive liberty \- Wikipedia, https://en.wikipedia.org/wiki/Cognitive\_liberty
15. Mind Control in the Digital Age (Mind Shaping and Cognitive Liberty) \- a Podcast, https://www.robsoncrim.com/single-post/mind-control-in-the-digital-age-mind-shaping-and-cognitive-liberty-a-podcast
16. Bublitz Draft My mind is mine Cognitive Liberty as a Legal Concept 2013 \- Antonio Casella, https://www.antoniocasella.eu/dnlaw/Bublitz\_2013.pdf
17. The European Union on Cognitive Liberty within New Technologies, https://www.humanrightsresearch.org/post/the-european-union-on-cognitive-liberty-within-new-technologies
18. The Battle for Cognitive Liberty in the Age of Corporate AI | TechPolicy.Press, https://www.techpolicy.press/the-battle-for-cognitive-liberty-in-the-age-of-corporate-ai/
19. LlmWikis.org \- LLM Wiki Handbook for AI Knowledge Bases, https://llmwikis.org/
20. JustAnIota Compact AI Messaging: ɩ.com, https://xn--8na.com/
21. Producibility Studies Using UAl for Low Enrichment Fuel Platesx \- INIS-IAEA, https://inis.iaea.org/collection/NCLCollectionStore/\_Public/36/023/36023723.pdf
22. Contact \- aRuntime.com, https://aruntime.com/contact/
23. 'An experience of meaning': A 20-year prospective analysis of delusional realities in schizophrenia and affective psychoses \- Advocate Health, https://institutionalrepository.aah.org/allother/311/
24. Cherise Rosen PhD \- Department of Psychiatry, https://www.psych.uic.edu/profile/cherise-rosen
25. Jasper Feyaerts \- Loop, https://loop.frontiersin.org/people/198786/editorial
26. 'An experience of meaning': A 20-year prospective analysis of delusional realities in schizophrenia and affective psychoses \- PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC9388349/
27. 'An experience of meaning': A 20-year prospective analysis of delusional realities in schizophrenia and affective psychoses \- ResearchGate, https://www.researchgate.net/publication/362478288\'An\experience\of\meaning'\A\20-year\prospective\analysis\of\delusional\realities\in\schizophrenia\and\affective\_psychoses
28. Helen Harrow's research works | University of Illinois Chicago and other places, https://www.researchgate.net/scientific-contributions/Helen-Harrow-2201738647
29. 'An experience of meaning': A 20-year prospective analysis of delusional realities in schizophrenia and affective psychoses \- Frontiers, https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2022.940124/full
30. Listening to Schneiderian Voices: A Novel Phenomenological Analysis \- PMC \- NIH, https://pmc.ncbi.nlm.nih.gov/articles/PMC4990463/
31. (A) Probability of experiencing a specific delusional construct in... \- ResearchGate, https://www.researchgate.net/figure/A-Probability-of-experiencing-a-specific-delusional-construct-in-individuals-with\fig2\_362478288
32. Illusory social agents within and beyond voices: A computational linguistics analysis of the experience of psychosis \- PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC9290020/
33. Who Are People with Psychosis Delusional about? A Study of Social Agents in the Phenomenology of Delusions \- PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC12867498/
34. Zsofia Demjen | Publications | University College London \- UCL Profiles, https://profiles.ucl.ac.uk/60290-zsofia-demjen/publications
35. The psychometric properties of the DAIMON Scale, a translation from Spanish to English: An instrument to measure the relationship with and between voices \- PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC7518389/
36. Self, Voices and Embodiment: A Phenomenological Analysis \- ResearchGate, https://www.researchgate.net/publication/301671794\Self\Voices\and\Embodiment\A\Phenomenological\_Analysis
37. Psychosis Research Program \- Department of Psychiatry, https://www.psych.uic.edu/research/psychosis-research-program
38. Psychosis Research Program Our Research | Department of Psychiatry | University of Illinois College of Medicine, https://www.psych.uic.edu/research/psychosis-research-program/our-research
39. Cherise ROSEN | Doctor of Philosophy | University of Illinois Chicago, Chicago | UIC | Department of Psychiatry (Chicago) | Research profile \- ResearchGate, https://www.researchgate.net/profile/Cherise-Rosen
40. The role of inner speech on the association between childhood adversity and 'hearing voices' \- PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC10731775/
41. Collaboration Award 2024 | University of Cambridge, https://www.cam.ac.uk/public-engagement/cambridge-awards/2024/collaboration-award
42. Understanding Senua's psychosis in Hellblade \- Journal of Geek Studies, https://jgeekstudies.org/2024/07/17/understanding-senuas-psychosis-in-hellblade/
43. University of Cambridge Unit of Assessment: 4 Title of case study: Hellblade \- REF 2021, https://results2021.ref.ac.uk/impact/9aeba804-2f6a-4864-a874-d9394bd025ed/pdf
44. The Town of Light \- Steam Community, https://steamcommunity.com/app/433100/positivereviews/?l=english\&browsefilter=toprated
45. The Town of Light for Xbox One review: Explore the terrors of mental illness and abuse, https://www.windowscentral.com/town-light-review
46. Preview: The Town of Light (PS4 version). \- One of those film blogs, https://oneofthem1s.wordpress.com/2017/03/03/preview-the-town-of-light-ps4-version/
47. A Contemporary Take on Victorian Lunacy: Representations of the Asylum in the Neo-Victorian Video Game Alice: Madness Returns – Journal of Games Criticism, https://gamescriticism.org/2025/11/05/bispo-6-1/
48. Feinberg Clinical Trials \- Psychiatry and Behavioral Sciences \- Northwestern University, https://www.psychiatry.northwestern.edu/research/clinical-trials.html
49. Behavioral Health Services, Psychiatry Services | Northwestern Medicine, https://www.nm.org/conditions-and-care-areas/behavioral-health
50. Clinical and Interventional Research \- Psychiatry and Behavioral Sciences, https://www.psychiatry.northwestern.edu/research/areas/clinical-and-interventional/index.html
51. Psychiatry and Psychology Services \- Rush University Medical Center, https://www.rush.edu/services/psychiatry-and-psychology-services
52. Critical Play: Depression Quest \- The Mechanics of Magic, https://mechanicsofmagic.com/2026/05/29/57179/
53. Depression Quest, A text adventure game which is also a learning tool. \- Off-A \- Asexual Visibility and Education Network, https://www.asexuality.org/en/topic/84159-depression-quest-a-text-adventure-game-which-is-also-a-learning-tool/
54. Learning about depression by watching gaming videos: a case study on the potential of digital games for psychoeducation and destigmatization \- PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC12146388/
55. Playing with the Narrative of Mental Illness: Communication Beyond Serious Empathy Games | BID, https://bid.ub.edu/wp-content/uploads/2024/06/kreitler.pdf