The Cost of Authenticity in AI-Driven Film

Artificial intelligence (AI) refers to digital systems designed to perform complex tasks that emulate aspects of human cognition, including interpretation, anticipation, reasoning and communication. While AI has existed for decades, the emergence of generative AI in 2022 marked a significant shift in its accessibility and growing acceptance. Integrating tools such as ChatGPT, Gemini, and Meta AI into everyday digital platforms has expanded access to AI-powered automation, changing how people approach research, editing, transcription, and creative production. By 2026, AI adoption had become increasingly widespread across industries, with 88 per cent of organisations reportedly using AI in at least one area of their business operations. Within the arts, particularly visual arts and film, AI is increasingly being used for scripting, story development, animation and video editing.

Traditionally, post-production can be time-consuming. Analysing and reviewing recorded footage requires careful examination to identify and select the most effective moments within a scene. AI-powered tools can streamline parts of this process by automating aspects of footage analysis and editing. DeepMotion, for example, uses AI-powered body-tracking and animation technologies to capture human movement and translate it into digital animation. Recorded movement can be transformed into 3D animation, allowing creators to reproduce human-like movement within digital environments.

AI-powered software such as Adobe Premiere Pro, which incorporates Adobe Sensei technology, can also assist with aspects of post-production by identifying elements within footage and supporting functions such as image stabilisation. But as AI becomes more integrated into filmmaking, its role extends beyond efficiency. It is beginning to influence how images are created, how stories are developed and, ultimately, how audiences understand what they see on screen. This raises questions around creativity, authorship, originality and the role of the filmmaker.

Over the past decade, LCA! has become one of Lagos’ distinctive cultural platforms. Yet the most important thing we have built is not a programme or a reputation, but a place of encounter: a community of filmmakers and critics, students and collectors, artists and entrepreneurs, and audiences of every kind, gathered by a shared belief in the power of cinema to bring people into conversation with one another and with the world around them. LCA! has always understood cinema not simply as something to be watched, but as a reason for people to gather. With the rise of AI, it is increasingly important that technology supports, rather than diminishes, the cultural perspectives and traditions that shape the stories being told, from Yoruba and Igbo traditions to those of Nigeria’s more than 250 ethnic groups.

The debate is no longer entirely hypothetical. Niyi Akinmolayan’s Colours of Fire (2025) brings cutting-edge technology and AI-powered visuals into a fantasy world centred around Moremi (Osas Ighodaro), a young woman from the Red Tribe, within an ecosystem divided by colour. In the film, the rival Red and Blue tribes are on a quest to discover what makes their pigments unique, how the colours can be transferred onto clothing and how they can be sold to Westerners. To uncover these secrets, Akinbode (Uzor Arukwe), a Blue Tribe warrior, is sent to kill the creature believed to have been unleashed by the rival Red Tribe. However, he encounters Moremi instead and discovers that she is not the monster he was sent to hunt.

The film combines CGI and animation with its fantasy narrative, incorporating cartoon-like caricatures and imaginative visual effects. Efunsetan (Mercy Aigbe), the leader of the Red Tribe, appears in a striking red ensemble, with a cape that is transformed into a weapon used to attack and capture enemies. Akinmolayan describes the film’s visual approach as “Afrofusion.” While Colours of Fire is an ambitious project within Nigerian cinema, some reviewers have criticised its use of AI-driven visuals, arguing that the technology can at times overshadow other strengths of the film, including its costumes, performances, colour symbolism and choreography. Among the film’s most compelling elements is its engagement with African mythology, particularly through Yoruba references, including the names Moremi and Akinbode, as well as Benin iconography and imagery from Northern Nigeria.

This tension points to a larger question about what audiences consider cultural authenticity. Film has always relied on technology to create worlds beyond our physical limitations, from practical effects and animation to CGI and digital collage. AI is part of this continuing history of technological change, but its ability to generate images, voices and performances introduces a different relationship between the filmmaker, the image and the audience. When a scene can be generated rather than filmed, cinema can move beyond representation into new forms of perception. Performance can become digitally altered, blurring the boundaries between what is physically performed and what is digitally created. What does it mean for audiences to know how an image was produced? And does knowing its origins determine whether an image, performance or film can be considered culturally, experientially or emotionally authentic?

LCA! has always been more than a screening programme. It has been an atmosphere in which cinema exists alongside music, hospitality, live performance and conversation; a space in which culture is not consumed from a distance but inhabited. These elements are not incidental to the festival; they are part of its philosophy. Cinema reaches its fullest meaning when it becomes the centre of a shared cultural life. The boundaries between the film on the screen and the life happening around it have always been intentionally porous.

As AI becomes increasingly embedded in filmmaking, perhaps the question is not whether filmmakers should use it, but how they choose to use it. Technology can expand what is possible on screen, but the stories, perspectives and cultural experiences behind those images remain human. For a cinema culture as diverse as Nigeria’s, that distinction matters. The challenge is to embrace new tools without allowing the technology itself to become more important than the people, histories and cultures that give the images meaning.

Bibliography

Notes

Multimodal emotion recognition (MER) is the process of using artificial intelligence to identify human emotions by combining multiple data types, such as facial expressions, voice tones, text, and body gestures.

The Impact of Artificial Intelligence on the Production and Editing of Audiovisual Content