人工智能对影视制作及行业未来的意义

Technology, Media & Telecommunications PracticeWhat AI could mean for film and TV production andthe industry’s future Industry leaders are questioning how AI could change what content ismade and how it is produced. Our research indicates three potentialindustry outcomes beyond disrupting the content supply chain. by Jamie Vickers and Marc Brodhersonwith Alec Wrubel and Cléophée Bernard AI is already beginning to be deployedin some areas of the film and TV production process,1though thepotential magnitude of its long-term impact is still coming into focus. Our research and discussions withstudio executives, producers, and technology leaders suggest that uncertainty around AI extends beyondwhether and how it will change production to how those changes manifest throughout the content anddistribution ecosystem. While the technology’s limits, adoption trajectory, and potential scale of impactare yet to be determined, historical technological shifts and early use cases suggest AI could, over time,materially alter the industry’s structure and profit pools. As a result, industry leaders face practical questions about near-term operating choices and strategicquestions about what AI could mean for their businesses longer term.2Based on our recent experience,research, and discussions, AI’s expanding capabilities have prompted some leaders to begin to reevaluatetheir business strategies while recognizing they must also manage looming concerns about labor impacts,potential risks, and the nature of creativity. As Sean Bailey, an industry veteran and founder of B5 Studiosdescribed the challenge to us, AI may represent “a more significant platform shift than we have ever seenbefore in our industry.” To understand how AI could impact the overall video content industry, we interviewed over 20 medialeaders, including studio and production executives, talent agents, AI innovators, and academics; drewlearnings from our work with video content companies; and analyzed broader industry data and the historyof technology innovations in content production and distribution (see sidebar, “About the research”). The rise in AI comes at a moment when video industry players are already under immense pressure.Consumer attention is fragmented amid an abundance of content and finite viewing time, and attention isshifting away from traditional channels to streaming platforms and user-generated content (UGC). In theUnited States, for instance, daily viewing hours spent on linear TV3declined by 4 percent CAGR from 2022to 2024, while streaming4grew by 13 percent and social video platforms grew by 14 percent (Exhibit 1).Consumers are also changing how they watch, consuming video on mobile devices and increasingly usingtheir TVs to search for and view online videos, including user-generated content.5 At the same time, investment in content is leveling off. In the United States, which represents over half ofglobal spend,6original content spend is projected to decline by 2 percent per year as buyers turn to sportsrights and licensed programming, which can attract outsized audiences or cost less (Exhibit 2). Theseprojections do not include the potential impact of AI, which introduces significant uncertainty aroundcontent spending. Amid these shifts in supply and demand, leaders interviewed indicated AI has the potential to impactmany production processes,7as well as back-office operations.8New tools and early experimentation arealready demonstrating single-digit productivity improvement potential in some use cases, while also raisingimportant questions about intellectual property (IP), authenticity, and the future of creative work.9 Exhibit 1 Video viewing time is projected to plateau, based on market momentum,with streaming and social video capturing a larger share. US total video viewing hours by video type,billions of hours McKinsey & Company Exhibit 2 Total content spend in the United States is leveling off as entertainmentcompanies focus on profitability over growth. US content spend by type,¹ excluding sports rights,$ billion McKinsey & Company While the longer-term implications are not yet clear, this research seeks to identify some potential industryoutcomes, starting with already-emerging impacts on production workflows. It also explores how AI couldchange what content is created and who creates and distributes it over the longer term under varyingadoption scenarios. The possible outcomes discussed include the democratization of high-end content creation, acceleratingthe shift of consumer attention from professionally produced content to UGC platforms and smallerstudios, as well as entirely new forms of content and distribution, including more immersive, personalized,or participatory entertainment. Our report also identifies the potential redistribution of economic valueand the potential for a net increase in total content supply and demand if AI’s impact is similar to pastdisruptions from the rise of new production and dis