South Africa’s film industry has been making great strides in the tech space. From our special effects to scenic spaces, it has become a hot spot for innovative filmmaking opportunities.
In a different beat, content producers from Dragon Studios have joined the fray but not for the traditional efforts we would expect. The studio was selected for this year’s Neu Wave AI Film Festival at LA Tech Week for their movie, Mountain of Graves.
Commissioned by Buka Africa Game Lodge, Mountain of Graveswas who are the brains behind Dragon Studios. The film has been produced using AI filmmaking platforms and follows two brothers fighting a mining company on a mountain where their ancestors are buried as nature starts to fight back
We catch up with Marsh, speaking about the film, the use of AI and the effect it has on jobs.
How is an AI film made?
Making an AI film is not an automated one-click process; it closely mirrors traditional filmmaking, with software in many instances replacing physical sets, cameras and actors. The journey begins with pre-production, where creators collaborate with language models to develop the script, shot list and scene “beats”.
To ensure the cast looks the same from shot to shot, filmmakers use AI image generators to establish “visual anchors” like character designs, lighting palettes and environments before any motion is created.
The “shoot” happens inside AI video generators, which typically animate still reference images, such as photographs, into short clips, with careful oversight from directors, who act as digital curators to produce a usable “take”.
And the voice actors?
Voice actors are synthesised with AI speech tools and specialised software syncs the characters’ lips to the spoken lines, with digital sound generators building ambient audio, sound effects and musical scores. The final phase returns to standard post-production in traditional editing suites. The filmmaker cuts the short clips together into a narrative timeline, aligns the dialogue and music, and polishes the cut.
While the work of “filming” happens in the digital space, there’s still plenty of space for human involvement in terms of creativity and control.
Vertical films are on the rise thanks to platforms such as TikTok. What has played a role in making them resonate with audiences?
The surge of vertical cinema stems first from an ergonomic shift: smartphones are designed for an upright, one-handed grip. By filling the vertical canvas, we eliminate black bars and peripheral distractions, creating an immersive, closed-world experience directly in the palm of the viewer’s hand.
The vertical frame also changes the way we tell stories. It isolates the human subject and prioritises the face, posture and micro-expressions, mimicking the perspective of a private video call or a direct conversation. This gives character-driven drama, internal monologues and suspenseful thrillers an intensity that feels personal and unmediated.
Why was it important to tell a story such asMountain of Graves, andwhat did the world and South Africans need to understand about the local mining industry?
Mining and land rights remain sharply relevant in Africa today because the global transition to green energy and digital technologies has triggered a renewed scramble for Africa’s critical minerals.
Vast reserves of these critical minerals lie beneath ancestral lands. This has placed unprecedented commercial pressure on indigenous territories and compounded the threat of displacement.
At the same time, the fundamental tensions between customary law and statutory authority remain unresolved in much of the continent. When national mining codes treat subsoil assets as state property while ignoring communal tenure systems, local leaders are pushed into protracted legal, physical and environmental battles.
This way of filmmaking requires a lot of technical skills and innovative thinking. What are the lessons learnt from other AI filmmakers that you hope to incorporate in future productions?
At Dragon Studios, we’re particularly inspired by the success of Gossip Goblin — the storytelling studio led by creator Zack London. The main takeaway from GG is that compelling AI filmmaking requires strict human authorship rather than relying on fully automated generation.
Every shot begins as a carefully directed still in Midjourney, undergoes editing and cleanup, and is then animated. By intentionally locking in the composition before motion is introduced, the workflow proves that cinematic coherence comes from treating the AI as an instrument for executing a specific vision, not as an autonomous director.
There have been many concerns about AI and its environmental impact. Don’t you think it’s contradictory to create a film using this technology while telling a tale that specifically looks at the same topic in a different industry?
We share the concerns about environmental impact and are advocates for greater regulation of AI. However, there are also benefits, especially for Africa and the Global South.
Fundamentally, we believe that AI can serve as a powerful equalising force for African filmmakers, lowering the capital and infrastructural barriers that have historically limited independent productions.
AI may actually be a net job creator in the long run, though this new model will cause some disruption as it emerges
— Donovan Marsh
Generative tools allow creators to bypass prohibitive studio costs for visual development, concept art and high-end visual effects, and directors can pre-visualise intricate worlds, prototype complex sets and deploy automated post-production workflows — an opportunity they ordinarily wouldn’t have.
The advent of AI filmmaking allows us to realise bigger dreams than we believed possible, and we want to push as hard as we can to get deserving stories told at amazing production value levels so we can compete with any film industry on earth.
South Africa’s film industry is on the rise thanks to several production companies filming in the country and creating jobs. Doesn’t an AI film take away from the need for jobs, or does it create jobs the public may not be aware of?
The model pioneered by Gossip Goblin outlines an emerging “micro-studio” ecosystem that inverts the traditional Hollywood pipeline: instead of financing a huge production and paying heavily to acquire an audience later, lean creative teams build a dedicated following first through serialised, short-form releases on open platforms.
Human craft remains central, with creators steering bespoke visual aesthetics, writing original lore and hiring human voice actors and sound designers to ground the performance. By maintaining low overheads and owning their intellectual property outright, these micro-studios can iterate rapidly.
This parallel ecosystem functions much like creator-led independent labels, using internet distribution as a launchpad rather than the final destination. Once an AI studio establishes a captive digital audience, it can cross the chasm into theatrical runs, anthology features and premium streaming from a position of commercial strength.
The result is a diversified entertainment landscape where small, self-contained production houses produce feature-scale cinematic universes at a fraction of standard industry budgets, permanently altering how films are developed, greenlit and distributed.
In this way, AI may actually be a net job creator in the long run, though this new model will cause some disruption as it emerges.
