Indian cinema is the largest film industry in the world by output, producing thousands of films and an endless stream of digital content every year. It is also, right now, in the middle of a quiet technological transformation. Artificial intelligence has moved from tech demos into the actual production pipeline, and the people most affected are the ones whose craft was built on the assumption that cameras, lights, and lenses are the only tools. This article looks at how AI is reshaping Indian cinematography, what it means for film students pursuing MFA degrees, and how both the industry and the classroom need to adapt.
Where AI is entering the production pipeline
The most common mistake is to think of AI in film as a single tool. In practice, it is a set of capabilities entering at different stages of production, each with its own implications.
In pre-production, AI helps with concept art, look development, storyboards, and visual planning. A cinematographer can describe a scene and get dozens of visual references in minutes, test lighting directions without renting a location, and align the entire crew on a look before a single camera is booked.
In production, AI assists with real-time effects, camera planning, and even automated editing of dailies. The cinematographer's eye still matters, but the range of what can be tested and visualized in real time has expanded enormously.
In post-production, AI is the most established: color grading, VFX cleanup, de-aging, background replacement, and sound design all have mature AI-assisted workflows. This is where the cost savings are already real, and where most working professionals have already encountered the technology.
The pattern across all three stages is the same: AI does not replace the cinematographer's judgment, but it dramatically lowers the cost of exploration. Decisions that used to require expensive tests can now be made visually, quickly, and with more options on the table.
The new model landscape and what it can do
The generation side of the field has advanced quickly, and Indian filmmakers now have access to a range of models that can produce footage from text, images, or other video. Each model family has different strengths, and the practical skill is knowing which one to reach for.
Photorealistic model families, like the Flux series, excel at detail, lighting, and style consistency, which makes them useful for visual development and for shots that must match a photographic look. They are strong tools for pre-visualization and for generating backgrounds and plates that would otherwise require a location scout and a full shoot.
Other model families, such as Sora and Kling, bring advanced motion and physics understanding. They can generate complex sequences, camera movement, and interactions that feel physically plausible. For Indian productions that rely on spectacle, song sequences, and action set pieces, these models are becoming a fast way to prototype choreography and blocking.
The most practical workflow is hybrid: generate a keyframe image that establishes the character and the environment, then use video models to animate it. This gives the filmmaker control over the composition while letting the model handle the motion. Consistency, the classic weakness of AI video, is solved by anchoring every shot to the same reference images.
Consistency: the technical challenge that matters most
Indian commercial cinema is built on continuity. A song sequence shot over multiple days, a character appearing in ten scenes, a hero performing in different outfits and locations, all of it depends on the audience believing the world is one world. AI's tendency to drift, changing a face, a costume, or a background between shots, is the single biggest obstacle to professional use.
The solution is reference-driven production. Build a character sheet at the start of the project: a set of images that define the face, hair, wardrobe, and key props. Reuse that sheet in every generation. Describe characters and environments with identical wording across all prompts, because models treat different words as different things.
The same discipline applies to light. Indian cinema has a rich visual language of color and light, from the warm tones of a wedding scene to the neon of a city night. Decide the lighting language for each sequence and repeat it in every prompt. The audience will not notice consistency when it works, but they will instantly notice when it breaks.
AI-assisted direction and the changing role of the cinematographer
One of the more interesting developments is the rise of AI agents that act as assistant directors: systems that take a creative brief, break it into shots, suggest camera language, and generate storyboards. The cinematographer's relationship to these agents is the key question.
The healthy relationship is collaboration, not delegation. The agent proposes; the cinematographer disposes. A good cinematographer brings the visual memory of decades of Indian and world cinema, an understanding of what a scene needs emotionally, and the judgment to reject a technically perfect shot that serves no story purpose. No model has that context.
What the agents genuinely add is speed and breadth. They can generate dozens of framing options in minutes, test lighting scenarios without a crew, and translate a verbal brief into concrete camera directions that the rest of the team can follow. The cinematographer who uses them well spends less time on mechanical exploration and more time on the decisions that only humans can make.
This shifts the required skillset. The cinematographer of the near future still needs the fundamentals, light, composition, movement, but also needs prompt literacy, the ability to direct AI tools precisely, and the ability to evaluate generated material critically. The craft is not dying; it is gaining a new instrument.
What MFA programs must change
MFA programs in cinematography and film were designed for a world where the camera department was the center of the technical universe. That world is changing faster than most curricula. If the degree is to remain valuable, several things need to happen.
First, the curriculum needs a serious technical track. Understanding how generative models work, how to build effective prompts, how to manage references for consistency, and how to evaluate AI output should be as normal as learning exposure and lensing. This is not a separate elective for tech enthusiasts; it is core craft for the next decade.
Second, the curriculum needs ethical and legal literacy. AI raises questions about authorship, likeness rights, data provenance, and labor that students will face in their careers. A cinematographer who understands these issues can protect their own work and navigate client and talent relationships with confidence.
Third, the curriculum needs industry integration. The gap between what film schools teach and what production companies need has always existed, but it is widening as the industry adopts tools faster than academia. Internships, studio partnerships, and projects that use real production software with AI features give students the practical fluency that classrooms alone cannot provide.
Fourth, the curriculum needs to teach the human skills that AI does not erode: visual storytelling, collaboration, leadership on set, and the ability to make aesthetic decisions under pressure. These are the durable parts of the craft, and they become more valuable as technical work gets automated.
Creativity, ownership, and ethics in the AI era
The adoption of AI in filmmaking is not just a technical question; it is a professional and ethical one. Cinematographers and film students need to think about what they are willing to automate and what they want to protect.
The ownership question is immediate. If an AI generates a look based on your references and your prompts, who owns the result? What happens when a model is trained on footage that you shot? The answers are still being worked out in law and in contracts, and professionals need to read their agreements carefully and advocate for clear terms.
The likeness question is already live. Actors and public figures have seen their images generated without consent, and the industry is responding with rights frameworks. Cinematographers should understand these issues because they are the ones operating the tools on set and in post.
The labor question is the most emotionally charged. Some workflows will automate work that was previously done by junior crew, and the industry must decide how to retrain and redeploy people. The honest position is that AI will not eliminate the need for skilled people; it will change what those people do. The professionals who adapt, by adding AI fluency to their craft, will be the ones who shape the next wave of Indian cinema.
Custom models and new revenue paths
One of the most promising developments for working cinematographers is the ability to customize models: training a model on your own visual library, your own characters, or your own brand of look. This turns AI from a generic tool into a personal instrument.
A cinematographer who builds a custom model trained on their own work can produce consistent visuals that look like them, not like the average output of a generic model. This is a form of creative signature, and it has commercial value: studios pay for a recognizable look, and a model that reproduces that look on demand is an asset.
The same logic opens new revenue paths for MFA graduates. Instead of competing only for traditional crew positions, they can offer AI-assisted services: pre-visualization, look development, concept art, and custom model building. These are services that did not exist a few years ago and that the market is actively demanding.
The educational implication is clear: students should learn not just to use models, but to customize them. Model training, dataset curation, and evaluation are becoming professional skills, and they fit naturally alongside the traditional craft training of an MFA program.
Bridging the gap between academia and industry
The gap between film education and the production industry is a recurring complaint, and AI is making it impossible to ignore. The fix is not to turn film schools into bootcamps; it is to make the practical bridge explicit and continuous.
Practical training needs to include real production software, not just theory. Students should leave with a portfolio that includes AI-assisted work, because that is what employers and clients will want to see. A reel that only shows traditional work will read as outdated in a field that has already moved.
The bridge also needs to run in both directions. Working professionals need continuing education as tools evolve, and schools can provide that through workshops, residencies, and certificate programs. The relationship should be a loop, not a one-time transaction.
Finally, the industry should be honest about what it needs. If production houses want graduates who can direct AI tools, they should say so in job descriptions and collaborate with programs on curriculum. The dialogue between academia and industry is not a courtesy; it is infrastructure for the future of the craft.
Frequently asked questions
Will AI make cinematographers obsolete? No. It will automate the mechanical parts of the job, but the aesthetic judgment, the storytelling instinct, and the leadership on set remain human skills. The cinematographers at risk are those who refuse to learn the new tools, not those who embrace them.
Do MFA degrees still make sense? Yes, if the program modernizes. A degree that teaches traditional craft plus AI fluency, ethics, and industry practice is arguably more valuable than ever, because the bar for entry into the industry is rising.
What should a film student learn first? The fundamentals of light, composition, and storytelling remain the foundation. Add prompt literacy and reference management early, because they multiply the value of everything else you learn.
Is AI-generated footage acceptable in professional films? It is already being used in many productions for pre-visualization, backgrounds, effects, and even finished shots. The question is not whether it is acceptable, but how it is used, and that is an artistic and ethical decision for each production.
How can a working cinematographer start? Pick one small task that AI can help with, pre-visualization for your next project, and use it end to end. Learn the workflow on something real, and build from there.
The road ahead for Indian cinema
Indian cinema has always been an industry of adaptation: from black-and-white to color, from film to digital, from theatrical to streaming. AI is the next transition, and the pattern is familiar. The technology changes the tools, but the audience still wants stories, emotion, and spectacle, and the people who deliver those are still the ones who understand the craft.
The cinematographer who embraces AI does not abandon the camera; they add instruments. The MFA student who learns AI fluency does not betray the art; they prepare for the industry that is actually hiring. The future of Indian cinematography will be written by the people who combine the best of the tradition, the eye for light and composition, with the best of the new tools, the speed and reach of AI.
Start the transition now, in whatever form fits your situation. Learn one tool, complete one AI-assisted project, ask the hard questions about ownership and ethics. Every step you take today becomes part of your craft tomorrow, and in an industry this large and this hungry for stories, the people who adapt will not run out of work.
