Authors:Simon Mutunga M’erimba

Research Publication | EAUR

HUMAN VERSUS AI STORYTELLING ON MENTAL ILLNESS: CASE STUDY MARIA AND TRAPPED IN A HORROR FILM

Authors:

Simon Mutunga M’erimba

East African University Rwanda

Filmmaking and Film Production

Abstract

Narration has been a critical aspect of human culture, which has served as a means of communication, education, and artistic expression for decades (Boyd, 2010). After the recent evolution in artificial intelligence (AI), the creation of stories has evolved, and it is hard to determine human-created stories and those created by AI. This has raised critical questions about the differences between human-authored and AI-generated narratives, even on sensitive topics like mental illness. This study examines how mental illness is portrayed in two contrasting filmic narratives: Maria, a Kenyan drama series aired on Citizen TV which is written by humans, and Trapped in a Horror Film which illustrates psychosis produced using AI. This study is based in an interdisciplinary theoretical framework drawing on posthumanism, AI aesthetics, and narratology. The research analysed the storytelling techniques used in both films, including characterisation, plot development, visual style, and emotional depth, to determine how AI-generated content compares with traditional human storytelling on stories about mental health issues. Storytelling among humans is often inspired by lived experiences, cultural contexts, and deep emotional engagement (Bruner, 1991). On the other hand, AI-generated plots rely on data-driven algorithms that synthesise patterns from existing texts and media (McKee & Fryer, 2020). This study investigated whether AI can authentically represent the complexities of mental illness or whether its approach differs significantly from human narrations regarding accuracy, empathy, and artistic nuance. Maria showcases a rich human perspective on mental health struggles through captivating character arcs, while Trapped in a Horror Film utilises AI to portray psychosis in a visually stylised manner. This raises questions about the ethical and creative implications of AI in filmmaking. A comparison of these two films helped to explore the potential for a complementary relationship between AI and human creativity in film production. The findings reveal a significant divergence between human-authored and AI-generated narratives and mental illness. While human-authored narratives embed mental illness within cultural contexts and character development, AI-generated narratives embed mental illness within visual stylization, genre-driven spectacle, and symbolic abstraction. It was also evident that there is a significant tension between human experience and algorithmic patterns, suggesting that algorithmic media may amplify aesthetic dimensions of narrative forms at the expense of cultural specificity and ethical complexity. The study supports emerging critical scholarship on the implications of algorithms for the emerging film media and narrative landscape. The findings corroborate the argument that algorithmic film can produce coherent narrative forms, yet these forms may not be embedded with human experience, knowledge, and morality. The study supports Floridi et al.’s (2018) and Bender et al.’s (2021) arguments that algorithmic media can produce coherent narrative forms, yet these forms may not be embedded with human experience, knowledge, and morality.

Keywords

AI-generated narratives; human-authored narratives; mental illness representation; AI filmmaking; African cinema; AI aesthetics; narratology; ethical storytelling.

1. INTRODUCTION

↑ Back to Table of Contents

This section deals with the background to the study, statement of the problem, objectives, research questions and contributions and implications of the study.

1.1 Background to Storytelling and Mental Health in Film

↑ Back to Table of Contents

Storytelling has existed as a universal human  form of entertainment through which individuals and societies express their experience, identity, and psychological states. Different cultures, use narratives as both cognitive and emotional means to help understand the human success, suffering, resilience, and healing (Bruner, 1991). In psychology, narrative theory puts a lot of emphasis on personal stories therby impacting how people interpret mental illness, how they construct meaning, and how they negotiate the social sense of belonging (McAdams, 2001).

Universally, storytelling is specific to the culture of the indigenous community but conceptually interconnected with other cultures. Western cultures put more emphasis on the subjectivity of individuals and internal psychological conflict in each individual in story telling, while other cultures such as Africa, and Asia, emphasise on the intersection between social relationships, communal responsibility, and spirituality (Chaudhary, 2018). In the African context storytelling is rooted in oral traditions like folktales, proverbs, and rituals. In this sense stories are used to help discuss trauma, moral conflict, and social imbalance (Finnegan, 2012; Okpewho, 1992).

Philosophies such as Ubuntu by Mbiti, view people as communal and social rather than as individuals. The mental well-being of a person is therefore understood in a holistic manner, shaped by social harmony, ancestors, spirituality, and community cohesion (Mbiti, 1990; Metz, 2011). Film, is a modern extension of storytelling, that inherits both global and local narrative traditions and so it is an important avenue where mental health meanings are culturally produced.

Film speaks to our subconscious mind and appeals to the people’s senses than any other entertainmnet medium because of its ability to combine sound, narrative, performance, and cinematography. It has therefore become a powerful tool of edutainment in regard to mental illness. Film shapes how the public perceives mental illness, and influences how audiences perceive and deal with mental disorders such as depression, schizophrenia, trauma, and addiction (Gabbard & Gabbard, 1999; Pirkis et al., 2006).

In the Hollyhood industry, mental illness individualised, institutionalised, and medicalised to reflect dominant psychiatric discourses (Hyler, Gabbard, & Schneider, 1991). On the other hand films from Africa, Latin America, and Asia view psychological distress an interconection of socio-economic, political, and historical contexts like gender oppression, war, colonial trauma, poverty and displacement (Chakrabarty, 2000; Diawara, 1992).

African cinema, emphasises social realism, collective memory, and postcolonial critique. Mental illness is therefore portrayed not as an isolated malady but as a response to breakup of social relationships, spiritual polarity, or historical violence (Ukadike, 1994). Popular African cinemas, including Nollywood and East African film movements, have made mental health films and they are consumed by mass audiences, making film not only an entertainment medium but also a social intervention on pertinet issues in society (Gugler, 2016). Film now has become a global media that operates as a cultural mediator to shape public perceptions, challenge silence, and reduce stigma.

Although cinema is a powefrful educational tool in society it also carries significant ethical risks when portraying mental illness. Research has consistently shown that cinema’s portrayal of mental illness often relies on stereotypes in society such violence, unpredictability, or moral deviance. This reinforces public fear and social exclusion of the mentally ill (Wahl, 1995; Diefenbach & West, 2007).

These risks are manifest differently in different cultural contexts. In Western media, misrepresentation often stems from over-medicalisation or the dramatisation of psychiatric institutions whereas in Africa misrepresentation may arise by showing supernatural explanations such as witchcraft or possession (Ebewo, 2007; Tomaselli, 2013). While Africa cultures perceive mental illness as a diseases that is related to spirituality African media scholars argue that such perceptions may lead to stigimatisation of mental illness by turning it into perfomance, rather than an illness that requires empathy and understanding.

According to Diawara (1992); and Mignolo (2011), if postcolonial African film scholars adopt dominant Western cinematic stories about metal health to african contexts then they will affect how indigenous societies understand mental health and promote cognitive inequality because many Africans irrespective of their levels of education sometimes consult the traditional seers and witchdoctors in regards to mental health. Filmmakers should therefore be sensitive to cultures, narration styles, and experiential realities and use such cultural understanding to make films that films that accurately address mental health discourse, humane, and socially constructive manner to help change the negative perceptions of mental illness.

The evolution of artificial intelligence (AI) in creating narratives has introduced new modes of authorship and creativity in film and media production. Modern AI systems make use of machine learning, natural language processing, and generative modeling to produce narratives with film script structures, character arcs, dialogue, and audiovisual sequences by combining patterns from extensive datasets of existing media texts (Manovich, 2019; Boden, 2016). In film production, people use AI to write scripts, create perfomances with AI generated locations and characters, editing, and generative visual design, which has reshaped conventional film production processes.

For this study, AI-generated film is not treated as an autonomous creative subject but as a computational film produced through aggregated narrative conventions rather than lived experience. This distinction is important and crucial to the methodological design of this research. Trapped in a Horror Film is used to represent an AI-assisted or AI-generated narrative text while Maria is used as a human-authored drama series. The study seeks to establish two contrasting modes of film production to allow for systematic comparison of how mental illness is constructed through narrative intent, character psychology, and aesthetic choices.

There is a great challenge nowadays when it comes to contemporary media analysis. As AI technolgy advances it is becoming difficult to distinguish between human-authored and AI-generated films. Advancement in generative AI has made stories produced by AI to closely replicate human narrative coherence, emotional pacing, and genre conventions (Bender et al., 2021). As a result of this challenge, narrative qualities on a surface level alone cannot sufficiently be used to determine film authorship or narrative depth.

At a methodological level, this study tries to respond to this challenge by going beyond questions of narrative fluency to focusing on the indicators ansd aesthetics of qualitative storytelling. These includes a focus on character development, emotional connectivity, narrative motivation, ethical shooting, and cultural influences because film is to a large extent influenced by the experience of the directors. The study draws from narrative analysis and film semiotics, to examine how each case study constructs meaning around mental illness by the use of plot development, visual symbolism, performance, and point of view. The assumption guiding this approach is that while AI can replicate narrative form, it may differ significantly from human storytelling in its capacity for psychological nuance, ethical reflexivity, and culturally grounded representation.

1.2 Problem Statement

↑ Back to Table of Contents

In the recent years AI is being used extensively in film production. There is a very thin line between human made and AI generated films which has made it difficult for the audience to decipher the difference between films made by human beings and those made using AI. There has been very little research that has been done to examine how AI-generated films portray mental illness as compared to human written stories, in regards to emotional depth, empathy, and cultural context. Human made films in African media often address mental health issues in relation to social, family, and cultural contexts, while AI-generated films mostly prioritise visual perfomance and and use of algorithms from exisiting stories over moral and cultural storytelling (Floridi et al., 2018; Bender et al., 2021). This is a critical issue for filmmakers, audiences, and scholars who seek to understand the implications in aesthetic, ethical and social AI-assisted storytelling in film. This study there will seek to bridge this gap.

1.3 Research Objectives

↑ Back to Table of Contents

The study aims to:

1.      Examine the storytelling techniques employed in human-written and AI-generated films.

2.      Compare the representations of mental illness by human made films and AI-generated films

3.      Find out the effectiveness of AI-generated storytelling regarding sensitive topics such as mental health.

4.      Investigate opportunities that exist for collaboration between human and AI in film making.

1.4 Research Questions

↑ Back to Table of Contents

The study is guided by the following questions:

1.      What storytelling techniques are employed in human-written and AI-generated films?

2.      Does the representations of mental illness by human made films and AI-generated films differ?

3.      How effective is AI-generated story regarding sensitive topics such as mental health?

4.      What opportunities exist for collaboration between human and AI in film making?

1.5 Contribution and Implications

This paper contributes to ongoing discussions film in regards to the implications of AI in the film industry offering an empirical, and theoretical comparison of human and AI storytelling. The study advances the concept of synthetic phenomenology as an important tool for analysing AI-generated narratives and underscores the ethical stakes if cultural representation is delegated to algorithmic systems.

The study argues that narrative meaning collapses into pattern when human factor is removed from the creation of stories, the study proposes that human authorship should be maintained at the story creation level when addressing mental illness issues. This positions the paper at the intersection of film aesthetics, AI contribution, and mental health representation, making it relevant to interdisciplinary scholarship.

2. LITERATURE REVIEW

↑ Back to Table of Contents

This sections focuses on literature that is related to the topic of study. The focus will be on storytelling, representation of mental illness, Ethics, and Empathy, in Mental Health Storytelling, and Artificial Intelligence and Narrative Generation.

2.1 Narrative Theory, and Meaning-Making

↑ Back to Table of Contents

According to Bruner (1991) narratives are not merely representational but constitutive and so they shape how individuals interpret reality, identity, and experience. In film, narrative techniques such as character arcs, cause effect, and temporal sequencing affect how audiences engage emotionally and cognitively with cinematic texts (Bordwell Thompson, 2019).

The way people process trauma is very cosely linked to story telling. According to McAdams (2001) film narratives help people to organise experiences into coherent identities, because the audience tend to copy a lot from the stories they watch and listen to. This process is particularly relevant in the analysing of the representations of mental illness. Therefore storytelling is not only cultural but also a psychological practice. This fact makes analysis of film narratives narratives a suitable methodological approach to use to examine films that address themes on mental health.

In africa sorytelling is traditionally a communal practice involving oral perfomance and relational identity. Okpewho (1992) and Finnegan (2012) argue that African narratives focus more on collective meaning over individual interiority, this means that african story telling addresses personal struggles within the society’s social, spiritual, and historical contexts. This is one reason that Maria chosen for this study because its an African film and in the film mental health is completely embeded in family dynamics, social pressure, and cultural expectations rather than isolated pathology.

2.2 Representation of Mental Illness in Film

↑ Back to Table of Contents

Mental illness has been studied widely in different fields, including psychiatry, media studies, and cultural studies. Researchers in different disciplines agree that film has a significant role in shaping public perceptions on issues such as mental health, thus functioning as a primary source of information for the audience (Gabbard Gabbard, 1999; Pirkis et al., 2006). However, although some films try to show mental illness in a positive manner, most films rely on stereotypes and popular views that associate mental illness with danger, irrationality, or moral failure (Wahl, 1995).

Narratives in Western cinema on mental illness emphasise that victims should be diagnosed, treated, and confined for a period during which they are sick. This approach encourages the audience to follow the medical and institutional procedures once an individual is diagnosed with a mental disorder (Hyler et al., 1991). Although this approach is medically acceptable, there are some mental disorders for which such procedures may not be necessary, as is the case with addiction and mild stress. This is in line with Byrne (2009), who observes that presenting mental illness in such medical and institutional ways may contribute to the sick person feeling socially distanced and discriminated against, making the majority hide their conditions or refuse to accept that they are sick altogether. Some films from Western countries also tend to exaggerate symptoms of mental illness for dramatic purposes, thus reinforcing stigma and misunderstanding.

African film, on the other hand,d showcases mental illness differently by associating mental illness with socio-economic and historical contexts of the African society. Diawara (1992) and Ukadike (1994) argue that African films show psychological disorders as an aftermath of colonial trauma, poverty, gender inequality, and social displacement. In popular African Films, such as Nollywood and East African television dramas like Maria, mental illness is addressed in the context of family relationships, moral dilemmas, and societal realities (Gugler, 2016). This way of presententing Psychological distress also has its demerits, and so scholars like Ebewo (2007) warn that portraying mental illness as being caused by witchcraft or supernatural causes risks reinforcing stigma and stereotypes.

2.3 Ethics, Empathy, and Responsibility in Mental Health Storytelling

↑ Back to Table of Contents

Ethics play a crucial role in the discussion of mental illness representation in movies. Ethical storytelling has been argued to involve accuracy, empathy, and respect for the human dignity of those living with mental illness (Wahl, 1995).

Another ethical consideration in mental illness representation is the concept of empathy. Empathy has been cited as an important ethical consideration in mental illness representation. Storytelling that evokes empathy in the audience has been argued to be effective in reducing mental illness-related stigmas (Pirkis et al., 2006).

Communal ethics of the narrative is also key in storytelling. This has been argued by African media theorists, who cite the importance of ethics in storytelling as encapsulated in philosophical theories such as Ubuntu (Metz, 2011).

These ethics are important in the evaluation of human-created stories as well as stories created by AI.

2.4 Artificial Intelligence and Narrative Generation

↑ Back to Table of Contents

Modern research in AI-generated narratives continues to explore the growing involvement of machine learning systems in creative industries. For example, AI-generated narratives make use of probabilistic models that detect and combine patterns from vast data sets, allowing them to create coherent narratives, dialogue, and films (Manovich, 2019; Boden, 2016). However, scholars argue that such systems are devoid of intentionality, consciousness, and experience.

Bender et al. (2021) argue that narratives generated by AI can be coherent and meaningful but are merely “stochastic parroting” surface-level coherence. When it comes to narratives that explore emotional and mental health realities use of AI in generating mental health films may be of great concern. For instance, Floridi et al. (2018) highlight the ethical issues in AI-generated content, including bias reproduction and unaccountability.

In terms of film and mental health storytelling, these issues are of critical importance. For instance, AI systems can be trained on existing media content and reproduce historically stigmatizing representations of mental health without ethical consideration (Benjamin, 2019; Noble, 2018). This literature highlights the need for critical comparative analyses of AI-generated narratives and their differences from human-authored narratives in exploring mental health themes.

While there has been considerable research on the representation of mental illness in film and AI-generated storytelling as distinct fields of research, there has been a lack of research on the comparative analysis of human-authored and AI-generated stories within the same conceptual framework, especially in relation to the representation of mental illness in film. This study will seek to bridge this gap by analysing the human-authored Kenyan drama series Maria and the AI-generated story of psychosis, Trapped in a Horror Film.

2.5 Theoretical Framework

↑ Back to Table of Contents

This paper is guided by an interdisciplinary theoretical approach that draws on posthumanist theories, AI aesthetics, and narratology. This paper is also guided by a posthumanist perspective that draws on theories of non-human authorship and agency (Hayles, 1999; Braidotti, 2013). The use of AI for narrative creation and creativity is also a part of non-human authorship and agency. Posthumanist theories have also helped to create a critical discourse on human perceptions of machine creation and human subjectivity.

On the other hand, AI aesthetics is a perspective that draws on theories of creating cultural artefacts through pattern recognition and probabilistic synthesis (Manovich, 2019). From this perspective, Trapped in a Horror Film is a cultural artefact created through the aggregation of horror tropes instead of experiential knowledge. In this case, there is a notion of synthetic phenomenology because AI simulates experience without having any experiential knowledge.

In conclusion, narratology is a perspective that draws on a critical discourse in regard to the difference between narrative form and human and AI authors (Genette, 1980). In this case, Maria a human-written story that draws on causal progression and ethical positioning. On the contrary, Trapped in a Horror Film draws on accumulated data and repetition.

3. METHODOLOGY

↑ Back to Table of Contents

This section will focus on the research design, sampling techniques, data collection analytical framework, and methodoligical limitations, ethical considerations.

3.1 Research Design

↑ Back to Table of Contents

The current study uses a qualitative comparative case study approach to critically examine the narrativisation and mediation of mental illness within human-authored and AI-generated filmic texts. Unlike most studies on narratives that treat it as a medium of representation, this study treats storytelling as a culturally and technologically situated practice that produces, negotiates, and contests meanings of mental illness. This is why a qualitative interpretive approach is most appropriate, given that it allows for an exploration of narrative, ethical, and discursive construction of mental illness that transcends quantification (Creswell Poth, 2018).

The comparative case study approach is a methodological approach that allows for an exploration of the emergent forms of algorithmic cultural production, including traditional forms of human-authored narrative. Comparative case studies are particularly effective when used to examine media forms that are produced under a complex interplay of cultural, technological, and ideological conditions, as is the case with AI-generated storytelling (Yin, 2018). By comparing a human-authored African television drama, Maria, with an AI-generated filmic text, Trapped in a Horror Film, this study interrogates how different regimes of authorship recode narrative structures, ethical positions, and representational logics of mental illness.

Essentially, the methodology is grounded in a comparative textual analysis, which is informed by theories of film, cultural studies, and critical media studies. The choice of Maria Trapped in a Horror Film is an analytical choice, as both stories employ the trope of psychological disturbance, fear, and institutional encounter, but they were produced through fundamentally different modes of production: human intentionality and algorithmic production. Rather than examining the clinical accuracy of mental illness representation, this study focuses on the epistemology of narrative: how meaning is constituted, how subjectivity is articulated, and how the viewer is engaged with a regime of affect and interpretation.

In terms of methodology, this study combines analytical tools from film studies, media ethics, and critical AI studies. The scripts and narratives will be analyzed with regard to causality, repetition, symbolism, affective modulation, and ideological framing. The interdisciplinarity of this approach will allow for an analysis not only of representation but also of the conditions of possibility that underpin representation. By examining how the regime of narrative authority and ethical engagement is transformed when authorship moves from human consciousness to algorithmic processes, this study will situate AI-generated stories within a broader discussion of cultural production, technological mediation, and representation. In this way, this study will contribute to the critical discussion in film studies with regard to authorship, agency, and the relationship between technology and narrative meaning in the current regime of audiovisual culture.

3.2 Sampling Techniques

↑ Back to Table of Contents

The research design used is based on purposive sampling, whereby two samples have been selected based on the fact that they are information-rich and relevant for the research (Patton, 2015).

3.2.1 Maria

↑ Back to Table of Contents

Maria is a Kenyan human-written drama television series aired on Citizen TV in Kenya. The drama series has been chosen for this research based on its presentation of mental health issues through long character development, character-based storytelling, and social realism.

3.2.2 Trapped in a Horror Film

↑ Back to Table of Contents

The research has chosen the AI-generated short film ‘Trapped in a Horror Film,’ which visually and stylistically portrays psychosis through the horror film genre conventions. The research has chosen this sample in order to analyse the ways in which AI generates mental illness narratives through the synthesis of conventional patterns of the horror film genre. The research has chosen one human-authored narrative from Africa and one AI-generated narrative in order to compare the two based on its objective of analyzing whether AI-generated storytelling can engage with complex mental health issues in a meaningful manner.

3.3 Analytical Framework

↑ Back to Table of Contents

The research is guided by a multilayer analytical framework, which combines narrative analysis, visual analysis, and ethical evaluation.

3.3.1 Narrative Analysis

↑ Back to Table of Contents

Narrative analysis is the study of the formation of narrative and the production of meaning in narrative, including narrative plot development, characterization, and thematic coherence (Bordwell & Thompson, 2019). The main narrative analysis categories are: Characterization and psychological development; Narrative development and causality; Use of dialogue or symbol and Resolution or absence of resolution This analytical perspective allows for the evaluation of how mental illness is represented as a lived experience, narrative, or spectacle.

3.3.2 Visual and Aesthetic Analysis

↑ Back to Table of Contents

Given the audiovisual nature of film, visual analysis is used to examine how cinematography, editing, mise-en-scène, and sound contribute to the representation of mental illness (Rose, 2016). Particular attention is paid to: Visual metaphors associated with mental distress; Genre conventions (e.g., realism versus horror) and Emotional tone created through lighting, color, and sound. This component is essential for evaluating how AI-generated imagery differs from human-directed visual storytelling.

3.4 Data Collection and Analytical Procedure

↑ Back to Table of Contents

The empirical material for this study primarily consists of the audiovisual texts Maria and Trapped in a Horror Film. These audiovisual texts are not understood as mere narratives but as culturally and technologically constructed spaces where meaning related to mental health is discursively produced and negotiated. The methodical approach to analysis can be understood as a systematic and reflexive process, which seeks to balance rigour and sensitivity to the aesthetic, ethical, and ideological aspects of representation.

The methodical approach to analysis began by watching the films several times. This was necessary to enable a deeper engagement with narrative structure, visual aesthetics, emotional tone, and representational strategies of each film text. This approach helped to identify patterns and tensions within each text, as well as allowing for the construction of an interpretive orientation to each text.

Following this initial level of immersion, a scene-by-scene granular level of analysis was undertaken, with specific attention given to those scenes where mental illness is highlighted as an experience, discourse, and/or institutional construct. The scenes were analysed in terms of their narrative, visual, and ethical aspects. The results of this analysis were then coded in terms of thematic frameworks based upon extant research in film and media theory and critical mental health. The coding of these results was undertaken in such a way as to be responsive to the emergent properties of the texts under study. This allowed for an ongoing dialogue between theory and empirical evidence.

As part of this study, a cross-case comparative analysis was undertaken in order to compare and contrast the representational logic of human-authored and AI-authored narratives. This allowed for an examination of both convergences and divergences in terms of their representation of mental illness. This level of analysis allowed for an examination of the ways in which different authorial logics impact upon the representation of mental illness in terms of narrative, affect, and ethical considerations.

As part of this level of analysis, an iterative and reflexive methodological approach was maintained in such a way as to continuously refine interpretive insights in response to repeated engagement with the data. This allowed for an inductive level of thematic emergence in response to both extant theory and the texts under study (Braun Clarke, 2006).

3.5 Ethical Considerations

↑ Back to Table of Contents

Films are very powerful when it comes to shaping the opinions, emotions, and stigma of society, so ethical considerations are key in this study Irrespective of the fact that the study did not use human participants in any capacity, it deals with cultural film texts that portray mental illness, making it a highly ethical study that calls for responsible analysis and interpretation of the film texts. The research paradigm that this study followed is based on respect, accountability, and reflexivity to ensure that any form of analysis or interpretation that was done did not fall into any form of stereotypical thinking or understanding of mental disorder. The ethical impact of AI in film narratives is of significant importance, especially in stories that have to deal with such complex and sensitive aspects of human life. In film, ethical representation is important to ensure accuracy, empathy, and social implications of any narrative representation of reality (Wahl, 1995; Floridi et al., 2018). However, it is important to note that AI does not have any moral or empathetic basis for generating stories or narratives, only using statistical patterns to generate content, which can have negative implications for stereotypical thinking, stereotypical representation, and aesthetic representation of mental illness (Benjamin, 2019; Bender et al., 2021).

Ethical analysis is an integral part of the comparative case study method employed here. Maria’s representation of mental health struggles is analyzed in relation to the development of the character, the social setting, and the narrative trajectory influenced by human authorship and cultural emotional epistemology. Trapped in a Horror Film, on the other hand, is analyzed based on the reliance of the narrative on genre, visual aesthetics, and abstraction in the representation of mental health struggles. The comparative method engages with the ethical implications of each narrative for the spectator, the way each narrative places the audience in an ethical relationship to the representation of mental health struggles, and the way each narrative constructs the audience’s relationship to the struggles of the character, such as the audience’s ability to empathize with the struggles of the character or the way the narrative constructs the audience as a consumer of the mental illness represented in the narrative.

Cultural sensitivity is another critical aspect of the analysis of the narratives. In analyzing the African media narratives, the research does not adopt the Western psychiatric perspective but instead situates the representation of mental illness within the social, historical, and cultural context of the society. African media theory and community ethical standards, along with the cultural philosophy of Ubuntu, are employed to analyze the narrative of Maria, focusing on the concept of relational identity, interconnectedness, and communal responsibility (Metz, 2011). On the other hand, AI ethics is employed to analyze Trapped in a Horror Film, focusing on the tensions between technological innovation and cultural sensitivity (Floridi et al., 2018).

Ethical perspective, in turn, is used in the evaluation of significant aspects of mental health representation in the two case studies. These include the balance between accuracy and exaggeration, humanisation and dehumanisation, and the level of cultural contextualisation in narrative structures (Wahl, 1995). The inclusion of ethical responsibility as a major analytical criterion in the research aligns the methodological orientation with international debates on ethical AI and representation in the creative industry. The research, therefore, seeks to contribute to the discourse that representation in AI-generated stories should not only be evaluated on the basis of competence but also on the potential and limitations in engaging with issues that require human understanding and ethical responsibility.

3.6 Methodological Limitations

↑ Back to Table of Contents

This research recognizes a range of methodological and epistemological limitations inherent in comparative qualitative research and in the analysis of culturally produced texts created through algorithmic processes. First, it is based on two case studies purposively selected for examination, and therefore it can only propose a limited scope of generalization to other textual and cultural sites of investigation. However, qualitative case study research prioritizes analytic generalization rather than statistical generalization, in which theory development is grounded in intensive examination of specific cases (Yin, 2018, p. 15). In this context, the scope of textual examination in this research allows for a rich exploration of narrative form, representation, and ethics in both human and artificial intelligence authorship.

A second limitation of this research lies in the difficulties of interpretation in relation to artificial intelligence authorship. The evaluation of Trapped in a Horror Film is also affected by a lack of transparency in terms of the data sets and processes used in creating it. This is a limitation inherent in artificial intelligence authorship in general, which reflects what scholars have termed the “black box” of algorithmic systems in which the logic of processes used in computation is not fully accessible to users or evaluators (Pasquale, 2015; Burrell, 2016). In this context, it is also important to note that algorithmic black boxes are not only a technological limitation but also a socio-political one in which meaning, agency, and responsibility are attributed to digital processes (Burrell, 2016).

Moreover, the idea of authorship in the context of AI-generated texts is itself subject to conceptual limits. Unlike human authors, whose agency is based on intentionality, consciousness, and morality, AI authors engage in statistical pattern recognition and, therefore, complicate the idea of authorship and narrative agency (Floridi et al., 2018; Bender et al., 2021). For instance, Bender et al. (2021) note that the narratives and texts generated by AI may have the appearance of textual coherence without any actual understanding of the subject. The above-mentioned limitations of the study are addressed by taking a reflexive and theory-informed approach that engages with the broader literature on critical media studies, film studies, and AI ethics. Rather than seeing the above-mentioned limitations of the study as a weakness, the study argues that the above-mentioned limits constitute the conditions of the research. The study's cautious approach, therefore, adds to the conceptual validity of the research and engages with the broader literature on the politics of representation in the context of emerging forms of algorithmic mediation and narrative agency (Couldry Hepp, 2017; Manovich, 2020).

4. ANALYSIS AND DISCUSSION

↑ Back to Table of Contents

This section presents an integrated comparative analysis of human-written and AI-generated narratives on mental illness using the film, Maria and Trapped in a Horror Film as case studies.nstead of treating the two texts as parallel and separate examples, this chapter analyses them concurently to examine the differences and similarties between human written and artificial intelligence made stories regarding narrative logic, character construction, emotional ethics, and epistemological assumptions.

The analysis shows that although both films engage with psychological distress, they do so using different narrative dimensions. Human stories constructs mental illness as lived experience which is embedded in social, emotional, and cultural contexts, whereas AI narratives construct mental illness as a patterned simulation assembled from already recognisable symptoms.

4.1 Narrative Structure and Plot Development

↑ Back to Table of Contents

The narrative structure of Maria follows pisodic format, with gradual character development that situates mental illness within family dynamics, social life, and personal struggles. For example Season 2 Episode 5 of Maria demonstrate how Luwi’s overreactive behaviour in response to Maria’s refusal to communicate is framed within ‘social interaction, portraying both cause and consequence while allowing for recovery arcs’ (Boyd, 2010; Bruner, 1991). Similarly, Victor’s verbal and nonverbal cues throughout Season 2 illustrate emotional strain and mental distress, ‘embedding illness within lived experience’ (Gabbard, 2001).

On the other hand, Trapped in a Horror Film presents a a perceptual loop narrative that is fragmented, and nonlinear. Trapped in Horror presents psychosis through fear, chaos, and visual distortion without any socio-cultural grounding. This augers well with AI’s algorithmic approach, ‘which prioritises symbolic intensity and spectacle over continuity and contextual meaning’ (Boden, 2016; Manovich, 2019).

Human storytelling in Maria puts emphasis on narrative continuity and relational meaning, while AI-generated narrative in Trapped in Horror Film amplifies ‘dramatic tension and visual spectacle at the expense of grounded context and social nuance’ (Bruner, 1991; Floridi et al., 2018).

4.2 Characterisation and Psychological Depth

↑ Back to Table of Contents

A central distinction between the two films lies in character construction. In Maria, the protagonist possesses psychological interiority. Thoughts, emotions, memories, and moral conflicts are used to show that mental illness results from complex human identity and interactions in society. The character does not exist in solitude but interacts with others, and thus, distress is shaped by human interaction through dialogue, misunderstanding, care, and resistance. This is evident in the case where Luwi’s mental illness is a result of a breakup with his girlfriend Brenda and the family’s insistence that he should Mary Sofia. Madam Vick is under mental pressure because of what is happening in her house, and while speaking to Sofia, she says, “I can’t believe this…my own house drives me crazy”

In Trapped in a Horror Film, the protagonist is not developed as a socially embedded individual but as a carrier of symptom clusters. Paranoia, surveillance delusions, referential thinking, and persecutory ideation dominate the narrative without biographical grounding. There is no mention of the character’s past or cause of his mental illness, relationships are distrusted or erased, and identity dissolves into pure perception. This results in a mechanised subject where consciousness is reduced to a system that processes threat signals rather than a human mind negotiating meaning.

Characters such as Luwi, Victor, and Vickie in Maria live a life that is intertwined, where mental illness is integrated into their identity, relationships, and the moral choices that each of them makes. Luwi’s depression is portrayed through wearing shabby clothes, his unkempt hair, and overreactiveness, which point to social withdrawal and distress (Gabbard, 2001; Byrne, 2000). Through violence, drunkenness, and suicidal ideation, Victor demonstrates stress, while on the other han,d Vickie is shown to be hysterical through exaggerated laughter and destructive acts like burning clothes in the compound while the whole family watches. These portrayals evoke empathy by allowing the viewer to be engaged with internal states (Bruner, 1991).

On the other hand, Trapped in Horror Film uses AI-generated characters and voice-over as symbols for externalising psychosis through exaggerated visual and voice-over narration, with limited psychological interiority (Manovich, 2019; Boden, 2016).

Mental illness in the AI-generated film emphasises symbolic intensity instead of contextual explanation. There is also use of surveillance imagery, distorted spatial continuity, and institutional environments such as hospitals, which are used as motifs to show loss of control and mental instability.

Unlike Maria, where mental illness is put in context with family conflict and emotional history, Trapped in a Horror Film portrays psychosis intense internal feelings and hallucinatory perception without any backstory or possible triggers. Human narratives invoke internal and relational circumstances, whereas AI prioritises the character's perceptions and the symptoms of psychosis, with a dramatic impact, detached from lived socio-cultural realities (Floridi et al., 2018).

4.3 Visual Style and Cinematic Language

Maria employs naturalistic cinematography, familiar domestic spaces, and visual realism to reinforce emotional credibility. Costume, facial expression, and body language are carefully used to signify psychological states, aligning with Goffman’s (1963) theories on social identity and stigma management.

Trapped in a Horror Film relies on stylized, distorted visuals, exaggerated lighting, colour and horror aesthetics, rendering mental illness visually spectacular rather than lived. AI’s visual output intensifies affective cues but risks reinforcing stereotypes and detaching the audience from empathetic engagement (Boden, 2016; Bruner, 1991).

African media theory highlights that storytelling in Kenyan television often relies on culturally legible cues, enabling audiences to interpret psychological states through socially contextualized behaviour (Finnegan, 2012; Ebewo, 2007).

4.4 Emotional Engagement and Empathy

↑ Back to Table of Contents

Human written narratives like Maria cultivate compassion and reflection, and frames mental illness as human and makes it socially situated. For example, Luwi’s anxious responses and Vickie’s destructive acts, ‘invite viewers to understand psychological pain through sustained narrative exposure’ (Bruner, 1991; Boyd, 2010). In this case emotional distress is made contextual, and distress is narratively understood even when the audience observe behavior that is disruptive or painful. The audience engangement is enhanced through empathy rather than merely observing.

The AI-generated story in Trapped in a Horror Film on the other hand emphasises fear and unease, thereby eliciting emotional reactions with limited empathetic identification. The focus on spectacle over the cognitave aligns with Bruner’s (1991) distinction between meaning-making and affective sensation in narratives. The AI-generated story, seems to construct mental illness through reproduction of the symptoms of Psychosis. The film strings together markers such as surveillance, conspiracy, institutional fear, and derealisation without offering any causes of the psychological illness or ethical framing. Mental illness therefore becomes an aesthetic system rather than a human condition. In this respect it can rightly be argued that human storytelling interprets mental illness, whereas AI storytelling reproduces its outward form.

4.5 Ethical Implications

↑ Back to Table of Contents

The ethical distinction between human and AI storytelling is evident. Maria demonstrates responsible representationcontextualizing mental illness within relational, social, and cultural frameworks. In contrast, AI-generated horror risks sensationalis, reinforcing stigma and dehumanizing conditions due to the absence of lived experience (Floridi et al., 2018; Bender et al., 2021).

In Maria culturally specific gestures and exaggerated costumes illustrate a delicate ethical balance: while culturally legible, hyperbolic portrayals can trivialize serious mental conditions (Diawara, 1992; Ebewo, 2007).

4.6 Human–AI Complementarity

↑ Back to Table of Contents

Findings suggest that AI can enhance visual metaphor and stylistic experimentation, while humans remain essential for ethical framing, empathy, and cultural specificity. A hybrid storytelling approach, combining AI-generated aesthetic innovation with human contextual and emotional grounding, is therefore recommended for sensitive subject matter like mental health (Floridi et al., 2018; Manovich, 2019; Boyd, 2010).

4.6.1 Integrated Comparative Summary

↑ Back to Table of Contents

Dimension Maria (Human Narrative) Trapped in Horror Film (AI Narrative)
Narrative structure Linear arc Perceptual loop
Characterization Interior, relational Symptom-based
Emotional logic Empathy-driven Fear-driven
Meaning-making Interpretive Algorithmic
Ethics Implicit moral responsibility Value-neutral
Mental illness portrayal Humanised Mechanised

4.7 Chapter Conclusion

↑ Back to Table of Contents

The comparative study of the two films demonstrates that stories generated by AI may imitate how metall illness is perceived by the sick person but  the stories cannot replicate the human meaning. The analysis done on Maria show that human storytelling prioritises contextualised storytelling, psychological depth, and cultural specificit the portrayal of mental illness. From Maria it is evident that character arcs involving mental health challenges are embedded with social, familial, and relational framework . The use of subtexts such as Luwi’s anxious gestures, Victor’s stress-induced drinking behaviour, and Vickie’s hysterical acts show how nonverbal cues and situational context are used to show psychological states, resonating with the emphasis of the African media theory on community-centered narratives and culturally legible communicatio(Finnegan, 2012; Diawara, 1992; Ebewo, 2007).

Although the film Trapped in a Horror uses horror conventions to show the inner world of the mentally sick and the stuggles that the psychotic goes through in their mind the story lacks it lacks the ethical intentionality, emotional continuity, and lived grounding present in Maria. Stories Generataed by AI prioritise spectacle, symbolic intensity, and algorithm-driven pattern , often detaching mental illness from relational and socio-cultural grounding (Boden, 2016; Manovich, 2019). It is true that AI can produce films that are visually striking and emotionally charge, but the stories lack ethical and cultural touch, thus such stories may lead to stigmatisation and dehumanisation of mental illness (Floridi et al., 2018; Bender et al., 2021).

The findings reveal that AI may function as a representational but not a consciousness-bearing storyteller. Stories by human authors narrate suffering through experiences shaped by the human context and care that the audience can associate with while AI stories narrate suffering as if it is a structure shaped by pattern. This epistemological gap has significant implications for the future use of AI in sensitive narrative domains such as mental health representation.

5. DISCUSSION, IMPLICATIONS AND RECOMMENDATIONS

↑ Back to Table of Contents

5.1 Introduction

↑ Back to Table of Contents

The comparative analysis of the narratives depicted in Maria, and Trapped in a Horror Film provides significant implications for the emerging landscape of mental illness narratives at the crossroads of human creativity and algorithmic media. The findings reveal that mental illness narratives are not just aesthetic and narrative tools but culturally specific forms that are shaped and mediated through regimes of authorship and algorithmic media. The analysis of these films, through the theoretical frameworks of global media and African media epistemology, provides significant implications for the emerging role of AI-generated media at the crossroads of narrative agency and spectatorship.

The findings reveal a significant divergence between human-authored and AI-generated narratives and mental illness. While human-authored narratives embed mental illness within cultural contexts and character development, AI-generated narratives embed mental illness within visual stylisation, genre-driven spectacle, and symbolic abstraction. The findings reveal a significant tension between human experience and algorithmic patterns, suggesting that algorithmic media may amplify aesthetic dimensions of narrative forms at the expense of cultural specificity and ethical complexity.

The study supports emerging critical scholarship on the implications of algorithms for the emerging media and narrative landscape. The findings corroborate the argument that algorithmic media can produce coherent narrative forms, yet these forms may not be embedded with human experience, knowledge, and morality. The study supports Floridi et al.’s (2018) and Bender et al.’s (2021) arguments that algorithmic media can produce coherent narrative forms, yet these forms may not be embedded with human experience, knowledge, and morality.

5.2 Ethical Implications for AI–Human Hybrid Storytelling

↑ Back to Table of Contents

The analysis thus underscores the pressing need for the development of ethical bases for hybrid storytelling formats, especially in domains of narration that involve human vulnerability, such as psychological conditions. Hybrid storytelling should thus be seen not just in terms of its potential for enhancing creativity through technology but also in terms of its socio-technical configuration in which ethics remains a human construct. This is evident in the need for culturally sensitive portrayals of psychological conditions in a manner that is empathetic and socially responsible, in accordance with the broader discussions in AI ethics on the need for a human-centered approach to AI (Floridi et al., 2018).

The analysis thus underscores the fact that culture remains a constitutive element of meaning in storytelling rather than a peripheral one. In fact, African media theories highlight the fact that gestures, language, humor, and the pacing of stories are all imbued with cultural values and experiences (Diawara, 1992; Finnegan, 2012). Thus, the AI-generated stories may be seen as culturally flattened or universalized portrayals of psychological conditions that require mediation through a human storyteller familiar with local epistemologies. This challenges the assumption of AI as a culturally unbiased creative tool in the production of hybrid stories since its data sets and ideological constructs are culturally situated.

The AI-generated stories thus foreground spectacle over empathy in the production of hybrid stories in terms of the prioritization of visual metaphor and emotional pacing in the story over the construction of narrative flow through which the audience can empathize with the psychological conditions of the characters in the story. This thus underscores the fact that ethics in hybrid storytelling goes beyond the technology itself to the need for a balance between technology and humanizing narration in the construction of hybrid stories in terms of themes such as psychological conditions in a manner that is transparent in terms of AI involvement in the storytelling process (Bender et al., 2021).

The study also suggests the productive possibilities of iterative co-creation between human storytellers and AI technologies. AI storytelling can be conceived as a dialogic process in which AI technologies enhance the possibilities for artists while human storytellers assume cultural and ethical accountability. This perspective on AI storytelling is consistent with cultural theories of digital creativity in which technological systems are conceived as co-actors in socio-cultural assemblages rather than as drivers of creativity (Boyd, 2010; Manovich, 2019).

5.3 Implications for Film Production, AI Development, and Mental Health Representation

↑ Back to Table of Contents

The implications of the study for African television and film production revolve around the application of AI storytelling technologies. While AI storytelling can enhance innovation and dynamism in film productions, the cultural and social content of storytelling depends on human authorship, particularly in the context of African cultures and societies (Diawara, 1992; Ebewo, 2007). The study suggests some of the possible risks and limitations associated with the application of AI storytelling technologies.

The implications of AI storytelling for film production and representation globally revolve around the innovation and dynamism that AI can bring to film production. AI storytelling can enhance dynamism in film productions while compromising human and ethical considerations in storytelling. This can be conceived as an evolution in media culture in which algorithms and AI technologies function as key mediators and drivers of information and cultural content production and dissemination.

With regard to mental health, the research has shown that hybrid forms of storytelling can help redress the balance between spectacle and psychological narrative. Within the context of such an ethical framework, it is possible for AI storytelling to cultivate empathy and critical thinking rather than fear, ridicule, and stigma. However, without proper mediation of the ethical and cultural implications of such forms of storytelling, the research has shown that the simplistic and sensationalistic portrayals of mental health and the stigmatization of those suffering from mental health issues may persist.

5.4 Recommendations for Scholarship, Creative Practice, and AI Governance

↑ Back to Table of Contents

The research has provided a number of recommendations that can guide scholarship, film practice, and AI research and development. First, the research recommends the use of hybrid forms of storytelling that bring together the benefits of algorithmic experimentation and human ethical and cultural judgment. Second, the research has recommended the development of institutionalised forms of ethical review for AI storytelling, particularly for sensitive issues such as mental health. Third, the research recommends the use of culturally diverse and representative data for training AI so that the risk of epistemological bias and cultural homogenisation is avoided. Fourth, the research recommends the need for cultivating the literacy of AI-mediated storytelling so that audiences can engage with such forms of storytelling and not assume that they are culturally authentic. Fifth, the research recommends further research into the psychological, cultural, and social impact of AI-mediated storytelling, particularly for the African media landscape, which has not been sufficiently explored in media studies.

Finally, the research recommends the use of culturally nuanced forms of subtext, body language, and relational narrative structures to convey psychological states without the risk of stigmatizing those suffering from mental health issues.

In conclusion, these implications and recommendations place AI-generated storytelling less as a matter of technological inevitability than as a site of cultural contestation in terms of its ethical and artistic potentialities. By exploring the interrelationship of authorship, technology, and culture, this research contributes to the larger debates in film and media studies about the potentialities of narrative agency, ethical representation, and the politics of algorithmic creativity in the digital era.

References

↑ Back to Table of Contents

Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021). On the dangers of          stochastic parrots: Can language models be too big? Proceedings of the ACM Conference on Fairness, Accountability, and Transparency, 610–623.  https://doi.org/10.1145/3442188.3445922

Boden, M. A. (2016). AI: Its nature and future. Oxford University Press.

Boyd, B. (2010). On the origin of stories: Evolution, cognition, and fiction. Amazon.

Bruner, J. (1991). The narrative construction of reality. Critical Inquiry, 18(1), 1–21.            https://www.academia.edu/9503047/The_Narrative_Construction_of_Reality

Byrne, P. (2000). Alcoholism and filmic representation: A psychological perspective. Journal      of Media Psychology, 12(2), 87–102.

Chakrabarty, D. (2000). Provincializing Europe: Postcolonial thought and historical         difference. Princeton University Press.

Chaudhary, N. (2018). Culture and the construction of the self. Integrative Psychological and      Behavioral Science, 52(1), 1–11. https://doi.org/10.1007/s12124-017-9410-1

Diawara, M. (1992). African cinema: Politics and culture. Indiana University Press.

Diefenbach, D. L., West, M. D. (2007). Television and attitudes toward mental health   issues. Journal of Broadcasting Electronic Media, 51(3), 450–472.            https://www.researchgate.net/publication/229500416_Television_and_attitudes_tow            d_mental_health_issues_Cultivation_analysis_and_the_third-person_effect

Ebewo, P. J. (2007). The emerging video film industry in Nigeria. Journal of African        Cinemas, 1(1), 45–57. https://www.researchgate.net/publication/294545438_The_emerging_video_film_industry_in_Nigeria_Challenges_and_prospects

Finnegan, R. (2012). Oral literature in Africa. Open Book Publishers.

Floridi, L., Cowls, J., Beltrametti, M., Chatila, R., Chazerand, P., Dignum, V., … Schafer, B.       (2018). AI4People—An ethical framework for a good AI society: Opportunities,    risks, principles, and recommendations. Minds and Machines, 28(4), 689–707.    https://doi.org/10.1007/s11023-018-9482-5

Gabbard, G. O., & Gabbard, K. (1999). Psychiatry and the cinema. American Psychiatric Publishing.

Gabbard, G. O. (2001). Psychiatry and the cinema. American Psychiatric Press.

Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. Prentice Hall.

Gugler, J. (2016). Film in the Middle East and Africa: Creative dissidence. University of  Texas Press.

Hyler, S. E., Gabbard, G. O., Schneider, I. (1991). Homicidal maniacs and narcissistic   parasites. Hospital Community Psychiatry, 42(10), 1044–1048.          https://doi.org/10.1176/ps.42.10.1044

Manovich, L. (2019). AI aesthetics. Strelka Press.

McAdams, D. P. (2001). The psychology of life stories. Review of General Psychology, 5(2),       100–122. https://doi.org/10.1037/1089-2680.5.2.100

McKee, R., Gerace, T. (2018). Storynomics: Story-driven marketing in the post-advertising      world. Amazon.

Metz, T. (2011). Ubuntu as a moral theory. African Human Rights Law Journal, 11(2),                 532–559.

Mignolo, W. D. (2011). The darker side of Western modernity. Duke University Press.

Mbiti, J. S. (1990). African religions and philosophy. Heinemann.

Noble, S. U. (2018). Algorithms of oppression: How search engines reinforce racism. NYU         Press.

Okpewho, I. (1992). African oral literature. Indiana University Press.

Pirkis, J., Blood, R. W., Francis, C., McCallum, K. (2006). On-screen portrayals of mental       illness. British Journal of Psychiatry, 188(6), 523–528.        https://pubmed.ncbi.nlm.nih.gov/16846952/

Showalter, E. (1985). The female malady: Women, madness and English culture, 1830–1980.       Virago.

Tomaselli, K. G. (2013). The cinema of apartheid. University of Chicago Press.

Ukadike, N. F. (1994). Black African cinema. University of California Press.

Wahl, O. F. (1995). Media madness: Public images of mental illness. Rutgers University  Press.

Filmography

↑ Back to Table of Contents

Trapped in a horror Film

Maria