S11.8 · Media, Entertainment & Creative
The sector's most AI-exposed segment, where generative AI is cutting VFX shot costs 30-70% and displacing the labor scarcity the services model was priced on.
The global visual-effects market is sized anywhere from $15-23B depending on scope (Technavio projects $15.24B of growth from 2025-2029) — a services slice inside the roughly $250B total global content-spend figure (Ampere Analysis). No segment in the sector is more AI-exposed. Major layoffs and studio closures, including a large post-production house's 2025 bankruptcy, are already occurring even as AI-in-VFX ranks among the fastest-growing categories in the space (Technavio projects $6.85B of growth from 2025-2029). The technology displacing the headcount and the product being sold into the industry are the same thing.
Demand originates wherever content is greenlit — principally the US and UK — while delivery work concentrates in hub markets: Canada, the UK and India. The two geographies are structurally separated, and neither side controls the other. The segment is fragmented, with DNEG, Framestore and ILM operating alongside thousands of smaller shops, and barriers to entry are low for boutiques equipped with AI tooling.
This is a pure work-for-hire layer sitting between production, which retains creative and commercial control, and finished deliverables. Post-production and VFX houses own no IP of their own and capture none of the upside if the finished title performs well commercially. Economically, that makes the segment the opposite pole of the sector from the rights-holding businesses: labor-based project revenue, thin and volatile margins, and a geography arbitraged around tax incentives rather than any durable competitive advantage. There is effectively zero annuity value anywhere in the segment, which leaves individual shops with little insulation against a slow production year.
Diversification into game-asset production and advertising production is the segment's main strategic move at present — defensive, aimed at reducing dependence on the cyclicality of film and television commissioning. Genuinely expansionary activity is scarce.
Generative AI is going straight at the core cost line — rotoscoping, compositing, de-aging and environment generation — cutting shot costs by an estimated 30-70%. And the moat being eroded is not a peripheral cost advantage. It is skilled-labor scarcity itself, the entire basis on which the traditional services model has been priced. Remove the scarcity premium and the pricing model built on it disappears with it.
Beneath the traditional post houses, a wholly new low-cost tier is forming: prompt-based, one-person AI VFX operations doing work that previously required an entire department. None of this is a forward-looking risk. The compression is already underway.