■ SAFE RISK ■ Sports & Entertainment
Live dance is a body in a room with other bodies, and that is the whole product. AI is reshaping the commercial and screen side of dance work faster than the stage side.
“AI generates choreography, but nobody buys tickets to watch a robot pirouette.”
Our AI replacement risk score — how we score jobs
A professional dancer's week is class every morning, rehearsal most of the day, performance at night, and injury management throughout. Beyond company contracts, most dancers stitch together income: commercial jobs, music videos, cruise and theme-park contracts, teaching, and increasingly social content that functions as an audition reel. Careers are short, bodies are the capital, and the difference between working and not is often a torn labrum.
The technology has arrived on the periphery first. Motion capture and pose-estimation tools now generate movement for games and animation without hiring dancers for every sequence, and generative video can produce dancing figures for advertising and short-form content that would previously have been a paid commercial booking. Choreographic software and AI tools suggest phrase variations and can adapt sequences to different body types. Notation and archival — historically a specialist skill — is being handled by video plus automated motion analysis. Casting and audition triage in commercial work already uses automated video screening.
The stage resists for structural reasons. Ticket buyers pay for co-presence, risk and virtuosity that can visibly fail, and the shared breath of a live audience. Robotic bodies are nowhere near the compliance and expressive range required, and would not sell tickets if they were. Teaching also holds: correcting alignment requires touch, eyes and a relationship. Our risk score of 21 reflects a career where the artistic core is safe but the commercial revenue that subsidises it — the ad shoot, the mocap gig, the corporate booking — is being eroded by synthetic media, which for many working dancers is the difference between sustainable and not.
Automatability: our editorial assessment of current and near-term AI capability
Commercial erosion is happening now: generative video is already replacing some advertising and short-form work that used to book dancers. Expect that to deepen through 2030 while stage and teaching demand hold. By 2040 the profession likely looks similar artistically but with a thinner commercial floor, making touring, teaching and direct-audience income more central than they already are.
Robots lack the compliance, weight-shift and expressive range that make dance legible as human effort, and even if they had it, the ticket-buying reason to attend would evaporate. Generated video dancers already exist and are used in advertising. That competes for commercial budgets, not for the audience that shows up to a theatre.
Through the commercial side. Synthetic video can produce a dancing figure for an ad or a short-form campaign without a booking, and motion-capture libraries reduce repeat sessions for games and animation. Those gigs subsidised many stage careers. Live performance and teaching fees are largely untouched, which shifts the income mix noticeably.
As realistic as it ever was, which is to say demanding and short. AI changes little about the fundamentals — competition, injury risk and thin pay early on. What has changed is that you should plan for teaching, choreography or adjacent work sooner, and treat audience-building as part of the job rather than self-promotion.
Yes, and few contracts address it well. Movement captured in one session can be retargeted onto other characters and reused across titles indefinitely. Ask what the data can be used for, whether it feeds model training, and whether there is any residual. Getting this right early matters more than the day rate.