AI in the Creative Industries: How Artists and Writers Are Adapting to the New Reality

The creative industries have entered a period of reckoning that few predicted would arrive this quickly. By mid-2026, artificial intelligence tools have moved from experimental novelties to embedded infrastructure across writing, visual art, music, film, and design. A survey published in April by McKinsey found that 62 percent of creative professionals now use AI tools daily, up from 18 percent just two years ago. The integration has been neither the utopian liberation that technology companies promised nor the apocalyptic displacement that artists feared. It has been, instead, a messy, uneven, and deeply consequential restructuring of how cultural work is conceived, produced, and valued. This report examines the state of that transformation across six creative sectors, drawing on industry data, legal filings, and interviews with practitioners navigating a landscape that shifts beneath their feet.

The Visual Arts: A Legal Battlefield

Nowhere has the collision between AI and human creativity been more visible than in the visual arts. Text-to-image systems — Midjourney, DALL-E 3, Stable Diffusion, and Adobe Firefly — have become standard tools in commercial illustration, advertising, and concept art. Their speed is undeniable: a process that once required a human illustrator hours or days can now be completed in seconds. The question that has consumed the industry is not whether these tools work, but whether they are legal.

The answer remains unsettled. In January 2024, the U.S. Copyright Office ruled that AI-generated images are not eligible for copyright protection unless a human author exercises sufficient creative control over the output. The decision created immediate practical problems: companies using AI-generated marketing materials cannot claim copyright over them, meaning competitors can freely copy the imagery. Getty Images responded by launching its own AI-generated image service in 2025, trained exclusively on content from its licensed library, guaranteeing commercial safety. "We built our AI on a foundation of legal clarity," said Craig Peters, Getty's CEO. "Every image our model produces traces back to content we have the right to use."

Hollywood's Uneasy Truce

The entertainment industry's relationship with AI was shaped decisively by the 2023 Writers Guild of America strike, which lasted 148 days and resulted in contractual protections limiting how studios could use AI in the writing process. The agreement stipulated that AI-generated material cannot receive writing credit, and that studios cannot require writers to use AI tools. Three years later, those protections have held — but the landscape around them has changed dramatically.

Every major studio now employs AI tools for tasks that fall outside the WGA contract's scope: script coverage, audience analysis, scheduling, budgeting, and marketing copy. Disney's internal AI system, developed in partnership with an undisclosed startup, generates detailed audience segmentation reports for every greenlit project. Netflix uses AI to analyze script pacing, identifying scenes where viewer engagement historically drops. These tools do not write scripts, but they shape the creative environment in which scripts are written. "The contract protects the act of writing," said David Goodman, former WGA president. "It doesn't protect the ecosystem in which writing happens. That ecosystem is changing faster than any contract can anticipate."

For performers, the threat takes a different form. Digital likeness technology — the ability to generate photorealistic video of a specific actor without their participation — has advanced to the point where background performers face existential questions. The 2023 SAG-AFTRA strike secured consent requirements for digital likeness use, but enforcement has proved difficult. An estimated 15 percent of background crowd scenes in 2025 theatrical releases used AI-generated extras, according to an analysis by the Visual Effects Society. "You can protect the A-list star's face," said casting director Jennifer Euston. "You cannot protect the anonymous person who used to earn a living standing in the background."

Adobe Firefly and the Licensed Content Model

Adobe's approach to generative AI has become the industry template — and the industry flashpoint. Firefly, Adobe's image generation tool launched in 2023, was trained exclusively on content from Adobe Stock and public domain works. The company positioned this as the ethical alternative to competitors whose training data included copyrighted images scraped from the internet without permission. The strategy has been commercially successful: Firefly is now integrated across Adobe's Creative Cloud suite, and more than 12 billion images have been generated using the tool since launch.

But the "ethical AI" positioning has drawn scrutiny. Artists whose work appears in Adobe Stock — many of them photographers and illustrators who uploaded content years before generative AI existed — argue they were not adequately informed that their work would train a machine that now competes with them. A class-action lawsuit filed in the Northern District of California in late 2025 alleges that Adobe's stock contributor agreements did not contemplate AI training and that the company's use of the material constitutes a breach of contract. Adobe has moved to dismiss the case, arguing that its terms of service explicitly permit use for "product improvement and machine learning."

The case matters because it will establish whether existing licensing agreements cover AI training — a question with implications far beyond Adobe. If the court rules that they do, companies with large content libraries will have a significant competitive advantage. If the court rules that they do not, the generative AI industry will face a reckoning over training data that could slow development for years. "Adobe is the test case for the entire creative economy," said Professor Pamela Samuelson, a copyright expert at UC Berkeley. "Whatever the court decides will reshape how AI companies acquire the data they need."

Music's New Economics

The music industry's encounter with AI has been characterized by rapid commercial adoption alongside fierce legal resistance. On the adoption side, AI-generated music has become a significant presence on streaming platforms. Spotify disclosed in its Q1 2026 earnings that AI-generated content accounts for 8 percent of total listening hours — a figure that represents billions of streams and raises fundamental questions about royalty distribution. When an AI-generated track is streamed, who receives the payment? The developer of the AI model? The platform? No one?

Spotify has attempted to address the question by requiring disclosure: uploads must be flagged as AI-generated, and royalty payments for undisclosed AI content can be withheld. The system is imperfect. "Spotify can't reliably detect AI-generated music," said Brian Message, partner at ATC Management, which represents Radiohead's Thom Yorke. "The technology to distinguish a human-produced track from an AI-produced one barely exists. You're asking a platform to enforce a rule it cannot verify."

The major labels have responded with both litigation and investment. Universal Music Group filed suit in 2024 against Anthropic, alleging that the company's AI models were trained on copyrighted lyrics. Simultaneously, Universal partnered with Google's DeepMind to develop AI tools that generate music under controlled licensing terms — a strategy that reflects the industry's pragmatic recognition that AI cannot be uninvented. "We're not fighting technology," said Universal CEO Lucian Grainge. "We're fighting for the right to control how our artists' work is used to build it."

The 62 Percent: What Daily AI Use Looks Like

The McKinsey finding that 62 percent of creative professionals use AI daily obscures as much as it reveals. The nature of that use varies enormously by profession, seniority, and geography. In graphic design, AI assists with background removal, color correction, and layout suggestions — tasks that were already partially automated and are now fully so. In copywriting, AI generates first drafts that human editors revise. In music production, AI handles mixing, mastering, and the creation of sample libraries. In architecture, AI optimizes structural calculations and generates design variations for client review.

The common thread is not replacement but acceleration. AI has compressed the time required to produce output, which has increased the volume of work expected from each individual. A mid-level graphic designer at a London agency described her experience: "In 2023, I produced three concepts per client meeting. In 2026, I produce twelve. The quality is higher because I have more time for refinement. But I'm doing the work of four people and being paid the same as one." This compression effect — more output per worker, fewer workers needed — has begun to show up in employment data. The Bureau of Labor Statistics reported in May 2026 that employment in "art and design occupations" declined 4.2 percent year over year, the steepest drop since the 2008 recession.

The creative workforce is adapting, but the adaptation is uneven. Established professionals with strong client relationships report that AI has increased their productivity and, in some cases, their income. Entry-level workers face a contracting job market where the tasks that once served as apprenticeships — stock illustration, template design, formulaic copywriting — have been automated. "The creative ladder is losing its lower rungs," said Dr. Christina Cacioppo, who studies labor markets at the London School of Economics. "The question is whether the profession can survive without a pathway in."

The Road Ahead

The legal battles currently underway — involving Getty Images, Adobe, Stability AI, Midjourney, and others — will not resolve the fundamental tension between AI capability and creative livelihood. Courts will establish boundaries around training data, copyright eligibility, and liability. Legislatures will write new rules. But the technology will continue to improve, and each improvement will test whatever boundaries have been drawn.

What is clear, as of mid-2026, is that the creative industries have passed the point of choosing whether to engage with AI. The engagement is here. The question now is on what terms. Artists and writers who have learned to direct AI tools — using them as instruments rather than replacements — report cautiously optimistic outcomes. Those who have resisted entirely face a market that increasingly views their resistance as inefficiency. The cultural production of the future will be neither purely human nor purely artificial. It will be a collaboration whose terms are still being negotiated, in courtrooms and studios and living rooms, by people who did not choose this transformation but must nonetheless live within it.