Artificial Intelligence in Theatre: Does it have a place?

Artificial Intelligence is transforming admin efficiency, research depth, accessibility and workflows, but its growing use in the theatre has sparked one of the biggest debates the industry has faced in decades. Theatre has embraced technological innovation for centuries, so why does generative AI feel fundamentally different from every advancement that came before?

If you'd asked theatre-makers in the 1800s whether electric lighting would destroy the industry, many would probably have said yes. Gas technicians feared redundancy, and audiences complained that electric light looked unnatural. Similar concerns surfaced when automated flying systems arrived, when digital lighting desks replaced rows of manual dimmers, and when projection mapping began appearing in major productions. Each innovation changed jobs, yet none attempted to replace the people responsible for creating the art itself. Lighting programmers emerged alongside computer-controlled desks. Video designers became an entirely new discipline. Sound designers embraced digital playback without composers becoming obsolete. Theatre evolved because technology expanded what artists could achieve rather than making artists unnecessary.

So, artificial intelligence deserves the same balanced conversation. Dismissing it entirely would ignore genuinely exciting opportunities. Assistive AI has already been used in theatres to improve accessibility through live captioning for deaf and hard-of-hearing audiences, generate multilingual subtitles, optimise rehearsal schedules around complicated company availability and monitor stage automation before faults become dangerous. The Royal Shakespeare Company’s research into AI as a tool explores exactly this relationship: technology that quietly removes repetitive administration so artists can spend more time creating. Nobody buys a theatre ticket because the rehearsal calendar was organised manually. If AI can reduce paperwork, improve access, and make productions safer without making anyone redundant, it has earned its place backstage.

The problem we all have is the use of generative AI. Automated scenery still needs a set designer. Computer-controlled lighting still requires someone to design every cue. However, generative AI attempts to replace creative decision-making itself. It can produce something that resembles a script, a generic composition of incidental music, or generate costume concepts and create promotional artwork within seconds. On paper, that sounds efficient. In practice, it misunderstands what creative departments actually do.

Take costume design. The audience eventually sees the beautiful results of months of hard work and sketches, but even those early drawings represent weeks of research, fittings, and collaboration. A costume designer studies historical references, sources fabrics that behave correctly under stage lighting, ensures performers can complete demanding choreography, works alongside lighting and sound departments, adapts designs during rehearsals, and solves dozens of practical problems audiences never notice. All generative AI can do is make an image by predicting what costume designs typically look like. It does not understand how that costume functions during a twelve-second quick change or why one fabric choice might create a microphone rustle while another remains silent. It creates illustrations rather than solutions.

Scriptwriting exposes the same weakness. Large language models can certainly write dialogue. The issue is that great playwrights don't simply arrange words into believable conversations. They draw upon lived experience, personal grief, political opinion, cultural observation and emotional vulnerability accumulated over decades. AI cannot remember childhood. It cannot experience heartbreak. It cannot develop an ethical viewpoint or discover something about itself during the writing process because there is no self to discover. It generates language by analysing existing human writing and predicting which word will come next in the sequence.

This leads directly into one of the biggest problems: where that material came from in the first place. Most generative AI systems have been trained on enormous collections of books, artwork, music, and designs created by real people, often without explicit permission or payment. Scenic artists, illustrators and costume designers are increasingly recognising their own personal flourishes beginning to appear within AI-generated images. Theatre has always borrowed from history, but borrowing traditionally involves interpretation. Generative AI risks becoming industrial-scale plagiarism.

Theatre already relies heavily on freelancers working contract to contract. Emerging designers build careers through early commissions. Young playwrights learn by having small productions staged. If producers replace concept artists with AI-generated images or commission fewer original scripts because software can produce something "good enough,” those early opportunities disappear. The industry may eventually create new jobs centred around AI ethics or workflow management, just as digital technology created new technical roles decades ago. But those positions need to be creative careers rather than replacing the people whose imagination built theatre in the first place with a bunch of software developers.

Most glaringly, audiences are instinctively aware of generative AI and have called it out at Something Rotten! and Thespians. Many low-budget and community productions are turning to generative AI to fill roles they don’t have the money or talent for, which is making people lazy. Audiences welcome technology that improves the experience, but become uncomfortable when it replaces craftsmanship. Part of theatre's appeal lies in knowing that every costume was constructed by skilled makers, every prop was chosen deliberately, and every line on the page began inside another person's imagination. Theatre is valuable precisely because it brings people together to share another person's interpretation of the world.

None of this means AI should be banished from the theatre. Quite the opposite. Assistive AI has enormous potential to improve accessibility, administration, archive management and technical reliability. These developments deserve investment because they strengthen productions without replacing the people responsible for making them. Generative AI needs to be quashed because it shifts technology from supporting creativity towards substituting it.

Theatre has survived every technological revolution by remembering what audiences actually come to see. They don't buy tickets because a computer generated an efficient solution. They come to witness craftsmanship, collaboration and imagination unfolding live before their eyes. Artificial intelligence can absolutely help theatre reach new heights, provided it remains in the wings rather than trying to take centre stage. After all, when it comes to creativity, the best prompt is still a person.