By
Stuart Kerr, Technology Correspondent,
LiveAIWire

Theatre has always been a collaborative art form, but the
collaborators were always human. That assumption is now being tested in
rehearsal rooms, design studios, and festival stages across Europe and North
America, where AI tools are moving from peripheral uses in marketing and
ticketing into the creative core of theatrical production. Scripts generated
by language models, stage visuals adapted in real time by AI systems reading
performer cues, audience sentiment analysed live to inform lighting and sound
decisions. These are no longer proposals for what AI might eventually do in
theatre. They are things happening now in documented productions, and they
are raising questions about authorship, credit, and artistic accountability
that the industry is only beginning to address.

A Stanford
University research project demonstrated how AI-generated
projections, trained on live audio cues and sentiment analysis of performer
delivery, adapted in real time to the emotional pacing of a production rather
than running on pre-programmed sequences. Design teams at experimental
companies are using diffusion models to develop moodboards, character
designs, and stage layout concepts during early production phases,
compressing processes that previously required multiple rounds of iteration
between directors and designers across several weeks. The efficiency gains
are real. What they cost creatively is a more contested
question.

The Scriptwriting Question

The most contested application of AI in theatre concerns writing.
Large language models have been trained to generate story arcs, develop
character backstories, and produce dialogue following specified rhythmic and
tonal parameters. Research
published in ResearchGate documents experimental productions in Europe
where groups used language models to write short plays in their entirety,
sometimes delivering AI-generated scripts without prior rehearsal
specifically to examine what emerges when performers must respond to dialogue
that no human constructed with them in mind.

These experiments reveal both the capability and the consistent
limitations of AI as a dramatic writer. A script is not simply a sequence of
words. It encodes a playwright’s understanding of what an audience needs to
feel and why, shaped by lived experience, dramatic tradition, and the
particular relationship between a writer and their moment in time that no
current language model possesses in any meaningful sense. What experimental
groups have consistently found is that AI-generated dialogue achieves surface
coherence and even moments of unexpected aptness while reliably missing the
subtler registers of human motivation, irony, and emotional risk that
distinguish significant theatrical writing from competent text
assembly.

What This Means for Theatre Practitioners

For directors, designers, and playwrights working today, AI tools
represent a new category of resource rather than a replacement for the
creative roles that define theatrical practice. The practical applications
that are already demonstrating value include rapid iteration of design
concepts, automated analysis of rehearsal recordings to identify pacing and
blocking patterns, audience data analysis that provides more granular
feedback than post-show questionnaires, and administrative automation of the
production management tasks that consume disproportionate time in
under-resourced theatre organisations.

The more contentious applications, AI-generated scripts,
AI-directed performances, AI systems with influence over live creative
decisions, raise questions that go beyond efficiency. Theatre is a live
encounter between human beings. Its power derives from the possibility of
failure, the actuality of presence, and the shared vulnerability of
performers and audience in real time. Whether AI can contribute to that
encounter or whether its presence fundamentally changes the nature of what is
happening is a question that different practitioners are answering
differently, and the diversity of those answers is itself a sign of genuine
artistic engagement with the question rather than reflexive adoption or
reflexive rejection.

Audience Analysis and Real-Time Adaptation

One of the less discussed but practically significant applications
of AI in theatre is audience analytics. Sentiment analysis tools applied to
audience sound data, including laughter, silence, restlessness, and applause
patterns, can provide feedback on moment-to-moment audience engagement that
no other method produces at equivalent granularity. This data can inform
post-performance editing decisions, programme curation, and in experimental
contexts the real-time adjustment of pacing and technical elements during
live performance.

The ethical dimensions of real-time audience analysis deserve more
discussion than they have received. Audiences attending a theatre performance
have not in most cases consented to having their acoustic and in some trials
visual responses monitored and analysed by AI systems. The use of this data
to shape future performances raises questions about the relationship between
audience and performance that go beyond the technical questions about what
the data reveals. Theatre has historically depended on the audience’s sense
that their response is genuinely received by human performers, not processed by
an algorithm. Whether AI audience analysis changes that relationship or
simply provides better information about it is a question that warrants
sustained attention from the industry.

Authorship, Credit, and the Institutional
Response

The Writers Guild of America’s negotiations over AI use in
screenwriting established a precedent that theatre unions and arts funding
bodies are now working to apply to theatrical contexts. The core questions
are whether AI-assisted or AI-generated work should be credited differently
from entirely human-authored work, whether writers who use AI tools retain
the same authorship rights as those who do not, and whether funding bodies
that support the creation of new theatrical work should develop specific
criteria for how AI assistance is disclosed.

These questions do not have settled answers, and the pace of AI
capability development means that the answers will need to be revisited
regularly rather than settled once. What the theatre industry can do is
ensure that the frameworks being developed are based on honest engagement
with what AI is actually doing in current productions rather than on either
the promotional claims of technology companies or the reflexive anxieties of
practitioners who have not engaged with the tools directly. For related
coverage of AI in creative industries, see our analysis of AI
and creative activism, the debates around whether
AI can be truly creative, and our coverage of how
generative AI learned to tell stories.

What the Next Five Years Look Like

The most plausible trajectory for AI in theatre over the next five
years is increasing integration in design and production support roles,
contested adoption in writing and directorial roles, and growing
institutional engagement with the governance questions that adoption raises.
The experimental productions that are testing AI in creative core roles today
will generate the evidence base that informs mainstream adoption decisions,
and the industry bodies, unions, and funding organisations that set the
frameworks for that adoption are already beginning to engage with those
questions.

The outcome that would most damage theatre as an art form is
uncritical adoption driven by cost pressure rather than creative possibility.
Theatre’s economic model is already under strain, and the temptation to use
AI to reduce the costs of script development, design iteration, and technical
production is real. If those cost reductions come at the expense of the
distinctively human creative investment that makes theatre different from
other art forms, the efficiency gains will be purchased at a cost that
audiences will eventually register even if they cannot name
it.

The more productive path is for the industry to engage with AI as
practitioners rather than as subjects of technological change, developing the
critical frameworks to distinguish useful applications from damaging ones and
building the institutional capacity to make those distinctions actively
rather than having them made by default by technology companies and budget
pressures. Theatre has survived and adapted through multiple technological
disruptions. The question is not whether it will survive AI but whether the
adaptation will be on terms that preserve what makes theatre worth
preserving.

About the Author

Stuart Kerr is Technology Correspondent at LiveAIWire, covering
artificial intelligence, cybersecurity, and the social impact of emerging
technology. He publishes daily at LiveAIWire.com.