Setting the Stage: Realism in Cinema’s DNA

From its very beginning, cinema has carried the burden and the promise of realism. When the Lumière brothers screened their first short films in 1895, audiences famously ducked as a train seemed to barrel toward them from the screen. The astonishment was not simply about movement or novelty — it was about a new kind of presence, an image that appeared to capture reality itself.

This foundational moment crystallized a tension that has haunted film theory ever since: is cinema defined by its ability to replicate the real, or by its power to construct illusions? Across the 20th century, the most influential film theorists framed realism as cinema’s core identity. Siegfried Kracauer, writing in the aftermath of World War II, described film as the “redemption of physical reality,” a medium whose greatness lay in its capacity to record the world in its contingency and detail. André Bazin, in his famous essay “The Ontology of the Photographic Image,” argued that the photographic basis of film anchored it in a unique relationship to reality, granting cinema a privileged link to truth.

For both Kracauer and Bazin, the photographic image was more than a representation: it was an indexical trace of the world, a record of “what has been.” This indexicality grounded cinema’s claim to realism. The moving image was, in their view, a window onto reality, even when stylized or fictionalized.

Of course, realism has never been a simple matter. Theories of montage, mise-en-scène, and genre demonstrate that realism is constructed through choices of framing, editing, and style. Yet the photographic base — the fact that light from the world left its imprint on celluloid — provided the foundation on which debates about realism could unfold. Even when cinema embraced fantasy or spectacle, it did so in dialogue with its anchoring in the real.

Digital technologies complicated this foundation, but they did not erase it. The shift from celluloid to digital cameras in the late 20th century sparked concern that cinema’s indexical bond was weakening. Still, digital images were usually tethered to captured reality: a digital camera, like a film camera, recorded light from the world, even if pixels replaced grain. Computer-generated imagery (CGI) expanded possibilities for spectacle, but it was largely folded into live-action footage, maintaining some continuity with indexical realism.

Artificial intelligence, however, marks a more profound rupture. When an AI generates an image, it does not begin with light bouncing off the world. Instead, it produces new images out of patterns in data, trained on vast datasets of prior images. The AI-generated image is not a record of “what has been,” but a synthesis of what could be made to appear. In this sense, AI challenges the very basis on which cinematic realism has traditionally been understood.

This raises the central provocation: if cinema’s identity has always been tied to realism — grounded in indexicality, framed by philosophical debate, and contested through style — what happens when AI severs or reconfigures that bond?

Enter AI: A Break in the Tradition

The arrival of AI-generated images represents a profound departure from previous shifts in film technology. Whereas the move from celluloid to digital cameras reconfigured the medium’s material base, it did not fundamentally sever the connection between the image and the world. A digital camera still captured photons, still inscribed traces of reality, still grounded its authority in what Bazin called the “imprint of the real.”

AI breaks this chain. Generative models such as diffusion networks or large language–vision architectures do not need to point a lens at the world to create an image. They generate synthetic images from probabilities learned across massive datasets of prior media. A face, a street, a battlefield, or a dreamscape can be conjured without ever existing before the lens. The AI image is not a trace of light but a statistical synthesis — an image that owes its existence to patterns in data rather than the presence of the world.

This rupture places AI images in a different ontological category than both celluloid and digital cinematography. They are not records of “what has been,” but artifacts of computational inference. For Kracauer or Bazin, this would amount to a radical dislocation. Realism, in their terms, depended on cinema’s privileged bond to reality; AI unmoors the image from that bond.

And yet, paradoxically, AI images are often consumed as if they were real. The rise of “deepfakes” demonstrates the unsettling plausibility of AI-generated faces and voices. Platforms such as Sora, Runway, Gemini, and MidJourney have begun producing video sequences that imitate cinematic conventions so persuasively that they blur into the familiar textures of recorded film. To a viewer, the AI image can feel no less “real” than the photographic one — even as its origins are fundamentally different.

This duality — ontological rupture and phenomenological continuity — is the heart of AI’s challenge to cinematic realism. On one hand, AI severs the image from the world; on the other, it invites audiences to perceive and respond to it as if it were continuous with cinematic traditions.

Filmmakers are already experimenting with these tensions. AI-generated actors populate short films. Scripts written in part by large language models are entering development. Music videos, advertisements, and experimental cinema increasingly lean on generative AI — using platforms such as Sora, Runway, Gemini, and MidJourney — to produce imagery that would be impossible or prohibitively expensive by other means. These works are not marginal curiosities; they are prototypes of an emerging media landscape in which AI is woven into production at every level.

The implications extend beyond aesthetics to labor and authorship. If AI can generate an actor’s likeness, who owns that performance? If scripts can be drafted by generative models, what becomes of screenwriting as a craft and profession? If realism itself can be manufactured statistically, what happens to the ethical responsibilities of filmmakers and studios toward audiences?

These questions underscore why AI cannot simply be slotted into the lineage of cinema as “just another tool.” Montage, sound, color, widescreen, and digital cinematography each expanded cinema’s expressive range, but they remained tethered to recorded reality. AI alters the very ground of that tether. It introduces a new mode of realism — one no longer bound to the indexical trace but to algorithmic synthesis.

To study this phenomenon requires more than technical analysis. It demands a rethinking of film theory’s central categories: index, image, realism, illusion, truth. It demands philosophical engagement with questions of perception, cognition, and ontology. And it demands ethical reflection on the responsibilities of representing reality in an era when reality itself can be generated anew.

This is the horizon of AI Cinematic Realism: a field that acknowledges the rupture introduced by AI, while exploring the continuities, perceptions, and debates that make its images feel real.

Philosophical Reframing of Realism with AI

If cinematic realism once rested on the indexical bond between image and world, then AI-generated imagery forces us to reconsider realism not as a property of the medium but as a phenomenon of experience. This is where philosophy becomes indispensable: it gives us the tools to ask not only what realism is but how it is produced, perceived, and trusted when the world no longer anchors the image.

Phenomenology and perception

Phenomenology has long provided a way to think about the moving image as an embodied experience. For Maurice Merleau-Ponty, perception is not passive reception but active engagement with the world through our bodies. The cinematic image, even when staged, resonates with us because it mimics and extends our perceptual engagement.

But what happens when the image no longer refers back to a captured reality? An AI-generated street scene, for example, may never have existed — but it can still “feel” real in the flow of cinematic experience. Our bodies and minds respond to it as continuous with perception, even if the world it depicts is synthetic. In this way, AI realism underscores phenomenology’s core claim: that realism is as much about perception and embodiment as about ontological fidelity.

Externalism and the extended mind

Cognitive philosophers like Andy Clark and David Chalmers argue that the mind extends into the world through tools, technologies, and representations. Cinema has always been part of this extension, shaping not only what we see but how we think and feel about seeing.

AI complicates this framework. When AI systems generate images, they are not merely extending human perception but reshaping it through algorithmic inference. The external scaffolding of thought is no longer simply representational (a photograph of something “out there”) but generative, producing realities that condition how we think about what is “out there” in the first place.

In this sense, AI realism demonstrates an intensified form of externalism: our mental and cultural lives are increasingly co-constructed with computational systems that generate, rather than simply mediate, the images we encounter.

Truth, illusion, and the index

Classic debates in film theory often circled around whether cinema revealed truth or manufactured illusion. Bazin insisted on cinema’s special affinity with reality because of its photographic basis. Soviet filmmakers like Eisenstein, by contrast, highlighted cinema’s power to manipulate and construct meaning through montage.

AI collapses this binary. Its images can look indexical while being wholly synthetic. A deepfake of a politician appears to bear the same “trace of the real” as a news broadcast, even though no such event occurred. The danger — and the fascination — lies in the way AI images blend the truthful aura of photography with the generative freedom of imagination.

For philosophers of realism, this is both a challenge and an opportunity. It forces us to move beyond indexicality as the guarantor of cinematic truth and toward a more nuanced understanding: realism as a dynamic negotiation between representation, perception, and belief.

The ethical dimension

Philosophy also reminds us that realism is never a neutral category. To call something “realistic” is to confer authority, to frame it as credible, as having weight in the world. AI-generated media therefore raise pressing ethical questions: What responsibilities do filmmakers and media producers have when deploying realism? What happens when audiences cannot easily distinguish between recorded and generated images? How do we safeguard the trust that realism has historically commanded without succumbing to manipulation or deception?

Toward AI Cinematic Realism

AI forces us to reconsider realism not as a static essence but as a practice — something constructed through images, technologies, perceptions, and cultural conventions. This reframing allows us to see continuity with earlier philosophical debates while also recognizing AI as a rupture.

Phenomenology helps us understand why synthetic images still feel real. Externalism and extended mind theory explain how AI systems reshape our cognition and cultural scaffolding. Ethical reflection pushes us to grapple with the stakes of representation when realism can be generated from data.

Together, these perspectives form the philosophical foundation of AI Cinematic Realism. They show that the “real” in AI media is neither purely illusion nor purely truth, but a negotiated space where perception, technology, and culture intersect.

Ethical, Labor, and Cultural Implications of AI Realism

If philosophy helps us understand how AI reframes realism, ethics reminds us why this matters. Realism is not a neutral aesthetic category; it carries cultural weight, shaping how audiences trust images, how industries value labor, and how societies negotiate truth. When AI begins generating realistic moving images, these stakes become immediate and pressing.

Likeness and labor

Perhaps the most visible ethical issue is the use of AI to simulate performers. Synthetic actors can be generated to play roles, mimic celebrities, or even extend the careers of deceased figures. From one perspective, this appears as creative possibility — a way to tell stories that might otherwise be impossible. From another, it destabilizes the foundation of acting as labor.

If a studio can generate a convincing performance without paying a human actor, what happens to the livelihoods of performers? Questions of ownership and consent become central: does an actor own their face, their voice, their mannerisms? Recent strikes and negotiations in Hollywood underscore the urgency of these questions, as unions seek to protect performers from having their likenesses used without control or compensation.

Authorship and originality

AI also unsettles the idea of authorship. A generative model trained on thousands of screenplays can produce dialogue that feels natural, even witty. But who is the author? The algorithm? The engineers who built it? The writers whose scripts were ingested into its training data?

Cinematic realism is often tied to authorial vision — the director’s eye, the screenwriter’s craft. AI-generated realism raises the specter of works without clear authors, assembled instead from vast cultural databases. The ethical stakes here are not only about credit and compensation but about the cultural meaning of originality in an age of synthetic media.

Truth and deception

Realism has long been associated with credibility. A documentary shot in cinéma vérité style carries a presumption of truthfulness, even if carefully edited. AI-generated images break this presumption. A deepfake or synthetic documentary can look as “real” as captured footage, yet contain wholly fabricated events.

This introduces new dangers for political manipulation, misinformation, and propaganda. If audiences cannot distinguish generated realism from recorded reality, trust in media as a whole may erode. At the same time, there are artistic opportunities: filmmakers can use AI realism deliberately to provoke reflection on what we believe and why. The line between ethical use and manipulative abuse will depend on transparency, context, and intention.

Cultural memory and historical representation

AI also affects how societies remember the past. Generating a historically “realistic” scene — say, an AI reconstruction of an event with no surviving footage — risks shaping cultural memory in ways that feel authentic but have no basis in recorded reality. While re-creations have always been part of historical storytelling, AI’s power to produce seamless, photorealistic images raises the stakes. Whose version of history will be generated? What ideological interests will guide these reconstructions?

Global inequities

AI’s labor implications extend beyond Hollywood. Much of the training data for AI models is scraped from global digital archives, often without consent from creators. The benefits of AI-generated realism may accrue to corporations and well-funded studios, while the costs — in terms of displaced labor, uncredited contributions, and cultural appropriation — fall on less powerful communities.

At the same time, AI tools may empower independent filmmakers, allowing them to produce work that rivals studio production values. The ethical challenge is to ensure that such empowerment does not come at the expense of exploitation elsewhere in the pipeline.

Ethical frameworks for AI cinematic realism

To respond to these challenges, we need frameworks that link aesthetics with ethics. Realism cannot be discussed only as a matter of style or perception. In the age of AI, it must also be understood as a matter of responsibility: Who creates images? Whose labor is used or displaced? Whose realities are represented, and whose are erased?

These questions demand collaboration across fields — philosophy, media studies, law, and labor rights — and across industries. AI Cinematic Realism is not only a theoretical category but also a site of ethical struggle. By naming this convergence, we can ensure that the study of realism in AI media remains attentive to the lived consequences of representation.

Toward AI Cinematic Realism as a Field

Across its history, cinema has been entangled with questions of realism: how faithfully the moving image can represent the world, how audiences perceive this fidelity, and what ethical or cultural weight such representations carry. From Kracauer’s notion of cinema as the redemption of physical reality to Bazin’s defense of the long take as a guarantor of cinematic truth, debates on realism have defined film theory. In the digital age, those debates shifted but persisted, as indexical capture remained at least symbolically tied to reality.

AI marks a break. It does not capture the world but generates new images statistically derived from training data. Yet paradoxically, these images often appear just as “real” as those recorded with a camera. This dual condition — the loss of indexicality alongside the persistence of perceptual realism — forces us to rethink the very foundations of cinematic realism.

Naming a new field

To meet this challenge, I have proposed the term AI Cinematic Realism. It signals a field of inquiry devoted to understanding how realism operates when the moving image is generated, not recorded, by machines. The term deliberately joins three domains: AI as the technological rupture, cinema as the historical site of realism debates, and realism as the conceptual anchor.

This field is not merely an extension of film theory into new territory. It is an interdisciplinary space that must draw from philosophy, media theory, cultural studies, ethics, and even computer science to make sense of its object. Just as early film theorists had to invent new categories to understand montage, sound, or color, we now need a vocabulary adequate to AI’s transformation of the moving image.

What AI Cinematic Realism studies

The field involves at least four overlapping areas:

1. Aesthetics: How do AI-generated images produce effects of realism? How do conventions of framing, editing, and narrative interact with statistical generation to persuade viewers of authenticity?
2. Philosophy: What does it mean for an image to be “real” when its referent never existed? How do phenomenology, theories of perception, and extended mind frameworks help us rethink realism in synthetic images?
3. Ethics: What responsibilities accompany the use of AI realism? How do issues of labor, authorship, consent, and truth claims reshape the ethics of cinematic practice?
4. Cultural and political stakes: How does AI realism affect collective memory, propaganda, and cultural identity? Whose realities are generated, and whose are erased?

By mapping these areas, AI Cinematic Realism positions itself as a coherent field of study and practice, not just a passing trend.

AI as continuity and rupture

It is important to stress that AI Cinematic Realism does not erase the history of realism debates but extends them. Many of the same questions that haunted early film theory — about illusion, perception, and truth — remain alive. Yet the conditions have changed. With AI, the ontological status of the image itself has shifted, forcing us to revisit these questions with new urgency.

In this sense, AI Cinematic Realism is both continuous with film theory and a rupture from it. It honors the legacy of cinematic realism while recognizing that the terrain has been irrevocably transformed.

From theory to practice

This field is not confined to theory. Filmmakers experimenting with AI-generated imagery are already producing works that embody these debates. Festivals, online platforms, and artistic communities are beginning to showcase similar works, many of them created with tools such as Sora, Runway, Gemini, and MidJourney — suggesting that AI Cinematic Realism is already a living practice. The ongoing AI Cinematic Realism series explores these developments in parallel, mapping how creative experiments and theoretical debates feed into one another.

At the same time, policy makers, educators, and activists are grappling with AI-generated realism in contexts beyond art — from misinformation to surveillance. By connecting these domains, the field can ensure that theoretical insights inform practice, and that practical challenges reshape theory.

A call for collaboration

The establishment of AI Cinematic Realism as a field will require collaboration. Scholars of film and media studies need to engage with technologists; philosophers must dialogue with artists; ethicists must listen to industry practitioners. Only through such cross-disciplinary exchange can we develop frameworks adequate to the challenges and opportunities AI presents. Initiatives like the AI Cinematic Realism series are one way of fostering these conversations across academic, artistic, and public spheres.

The urgency of AI Cinematic Realism

Realism has always been at the heart of cinema, but never before has it been so radically destabilized and reimagined. AI forces us to ask not only what realism is, but what it should be in a world where images can be convincingly generated without referents.

By naming and framing AI Cinematic Realism, we create a space for rigorous study, creative experimentation, and ethical reflection. This is not just an academic exercise. It is a necessary response to a cultural transformation already underway. The moving image — once the trace of the real — has become a site where the real and the synthetic intertwine. To understand and navigate this new terrain, we need AI Cinematic Realism as both a concept and a field.