While studios publicly fought artificial intelligence in 2023 labor negotiations, their actual hiring patterns tell a starkly different story
The Strike Settlement and Its Promises
In the autumn of 2023, Hollywood’s creative workforce achieved what had seemed unlikely just months earlier: contractual protections explicitly written to constrain how studios could deploy artificial intelligence. The Writers Guild of America concluded its five-month strike on September 27, 2023, with a Minimum Basic Agreement that prohibited studios from using generative AI to write original material independently and explicitly banned the use of writers’ work to train AI systems. Seven weeks later, on November 9, the Screen Actors Guild-American Federation of Television and Radio Artists (SAG-AFTRA) reached its own settlement, secured after 118 days of work stoppage, which established that performers’ digital likenesses could not be scanned, replicated, or reused without explicit, informed consent and separate compensation.
These agreements represented, by any historical measure, a decisive assertion of labor power in an industry accustomed to capital’s upper hand. Industry observers hailed them as watershed moments—the first time artificial intelligence had been substantially constrained by organized labor, and the template against which future technology-related labor disputes would be measured.
Yet by mid-2026, the operational reality inside major studios appeared to diverge significantly from the protections those contracts promised. While the formal agreements remained in place and unions continued public advocacy against AI threats to employment, internal hiring patterns revealed something more complicated: Hollywood’s largest employers were simultaneously conducting perhaps their most aggressive recruitment of artificial intelligence talent in the industry’s history.
A survey of job postings conducted by the Los Angeles Times in June 2026 found that more than one in ten job listings from major studios appeared connected to artificial intelligence. The analysis examined hundreds of postings and identified two primary categories of recruitment: positions designed to build AI tools and infrastructure, and roles intended to defend studio intellectual property against unauthorized AI use. This marked a substantial acceleration from baseline patterns. According to data compiled by the jobs platform Indeed, the share of AI-related job postings in the arts category had stood at roughly five percent in May 2025. By April and June 2026, that figure had doubled to approximately eleven percent—a growth rate substantially exceeding the overall increase in AI job postings across all industries, which rose from 2.8 percent to 5.5 percent in the same timeframe.
The Hiring Signal
Job listings in the Hollywood area reflected this pattern. At the time of this reporting, 182 data scientist positions were open in Hollywood specifically, according to Glassdoor listings from late September 2025. Another snapshot showed 1,776 artificial intelligence positions available across the Los Angeles region. Positions included roles as machine learning engineers, AI training specialists, applied AI scientists, and LLM integration specialists. The concentration of these openings contrasted with public studio messaging that had, just two years earlier, emphasized the existential threat AI posed to creative employment.
The AI Training Economy
A second, less visible hiring phenomenon emerged alongside direct employment. The Hollywood Reporter documented in June 2026 that numerous television and film writers had taken contract work training artificial intelligence systems through platforms designed specifically for that purpose. These platforms, including Mercor and Handshake, connected domain experts—creative professionals whose livelihoods had been eroded by structural changes in the entertainment industry—with companies seeking to improve machine learning models through a process known as Reinforcement Learning from Human Feedback (RLHF).
Mercor, founded in 2023 by three Thiel Fellows, had grown into a substantial operation by 2026. The platform raised $350 million in Series C funding during 2025, achieving a $10 billion valuation. Its business model focused on recruiting domain experts to evaluate AI model outputs and provide comparative feedback—work that demanded the kinds of creative judgment and professional knowledge that traditionally commanded premium compensation in Hollywood. Creative writers could earn up to $44 per hour through the platform’s job listings; music professionals with advanced degrees could earn up to $100 per hour. This represented meaningful income, particularly for a workforce facing sharp employment contraction in traditional roles.
The emergence of this labor market was not incidental to the broader hiring pattern. It represented a structural adaptation to the collapse of traditional employment opportunities. The relationship between the two phenomena—aggressive AI hiring by studios and the simultaneous availability of displaced creatives for AI training work—suggested an unintended but consequential feedback loop: studios needed domain experts to fine-tune AI systems; unemployed writers and creatives needed income; platforms mediated the transaction.
The Context: Four Years of Contraction
Neither the AI hiring surge nor the emergence of the AI training economy should be understood in isolation. Both occurred against the backdrop of sustained, severe employment contraction in traditional film and television production.
According to Bureau of Labor Statistics data analyzed by industry observers, Los Angeles County shed 41,000 film and television jobs over a three-year period. This represented a loss of approximately one-quarter of the region’s entertainment workforce. The decline was not limited to on-screen and writing roles. The collapse in production volume affected marketing firms, public relations shops, craft services providers, freelance editors, and the constellation of specialized suppliers that had historically sustained a stable ecosystem around major production. For the million individuals whose livelihoods depended on traditional film and television in Southern California alone—a figure that included not only the remaining 130,000 unemployed or underemployed actors, but also lawyers, agents, accountants, and financiers—the contraction represented a structural reordering of the labor market.
The contraction did not affect all creative trades equally. Available work for writers declined by forty-two percent between 2023 and 2024, according to data cited by Deadline in tracking industry employment patterns. This made the appearance of well-compensated AI training work particularly significant for that workforce segment—it arrived at a moment when traditional income sources had largely evaporated.
On-location filming in the Los Angeles area experienced its worst year on record in 2024, excluding only the pandemic year of 2020, according to comments from Paul Audley, president of FilmLA, a nonprofit organization that tracks filming permits and production activity. “Right now, we’ve just come out of the worst year on record, excluding Covid, for on-location filming,” Audley stated in spring 2025. As the first quarter of 2025 drew to a close, he observed that the trajectory was “doing even worse.”
The Layoff Sequence
Within this broader employment collapse, individual studio decisions became visible markers of the transition. In October 2025, Paramount announced a sweeping round of layoffs affecting more than 2,000 employees—ten percent of its workforce of approximately 20,000. The cuts occurred within months of the Trump administration’s approval of the Skydance-Paramount merger, and struck across both entertainment and administrative functions.
Warner Bros. Discovery had conducted similarly consequential workforce reductions. Between 2023 and 2024, successive waves of layoffs eliminated hundreds of positions across CNN, Turner, and HBO Max, according to accounts from analysts examining the company’s restructuring. In a single restructuring during 2024, Warner Bros. Discovery eliminated nearly 1,000 employees. The financial impact of these cascading reductions reflected larger structural challenges. The company had recorded net losses exceeding $20 billion cumulatively across 2021-2025, with particularly severe impacts in 2024 when losses approached $11.3 billion to $11.5 billion.
CNN, the news division within Warner Bros. Discovery, announced approximately 200 job cuts in January 2025 as part of a broader digital strategy pivot. The company simultaneously announced that its parent would invest $70 million in CNN’s digital operations—a pattern that repeated across the industry: contraction in employment, reallocation of investment toward emerging capabilities.
Understanding the Apparent Contradiction
The simultaneous occurrence of aggressive AI hiring and massive workforce reductions, coupled with the emergence of AI training work as a substitute for traditional employment, invites several possible interpretations. Some observers have characterized it as a deliberate strategy of labor replacement; others frame it as an unintended consequence of separate business decisions made by different divisions within diversified media conglomerates. Closer examination of available evidence suggests the dynamic may be more complex than either framing captures.
The most straightforward explanation is economic pressure. Hollywood’s traditional business model—built on recurring revenues from theatrical exhibition, television advertising, and cable subscription fees—has contracted across multiple fronts. Streaming platforms cannibalized theatrical audiences and disrupted television’s advertising-supported economics. Studios faced pressure from investors to demonstrate path-to-profitability amid declining legacy revenues. Under such pressure, cost reduction through automation and capability rationalization becomes a logical business response, independent of labor-relations negotiations.
This economic logic manifested visibly in production choices. Netflix, in 2025, debuted a film titled The Eternaut that featured an entirely AI-generated building demolition sequence. Industry veterans who had tracked artificial intelligence’s potential impact on creative employment since 2022 and 2023 reported that storyboard artists, sound designers, and visual effects specialists—initially skeptical of AI’s capacity to produce professional-quality work—had by 2025-2026 begun integrating AI tools into their professional practice. These were not studio mandates imposed from above in most cases; rather, individual craftspeople were adopting tools to remain economically viable within production workflows constrained by budgets.
A second, parallel pressure stemmed from intellectual property defense. Studios faced mounting legal challenges to their content from AI companies that had trained systems on copyrighted film, television, music, and creative writing without permission or compensation. Disney, NBC Universal, and DreamWorks filed a landmark lawsuit against Midjourney in June 2025 alleging mass copyright infringement for the use of the studios’ characters and artworks in training. Hollywood studios including Disney and Paramount began alleging in 2026 that ByteDance’s Seedance 2.0 video generator had engaged in “pervasive copyright infringement” by using their films as training material. These legal battles created demand for specialized legal and technical expertise—to defend intellectual property, understand AI architecture well enough to build protections, and navigate regulatory uncertainty. This demand was separate from, though operating in parallel with, the cost-reduction logic driving AI-assisted production.
The Question of Enforcement
The 2023 labor agreements, while substantive in their language, contained structural gaps that became apparent as studios implemented AI systems operationally. The SAG-AFTRA contract’s digital replica provisions, for example, limited remedies to “monetary damages” if a studio created a performer’s digital likeness without consent. The agreement contained no requirement that studios delete or cease using an actor’s digital data after paying damages. As union members and critics observed, this meant studios could theoretically pay damages once and continue deploying a performer’s likeness indefinitely. For stunt performers and specialists in physically dangerous work, the implications were particularly stark: studios could substitute AI replicas for human performers in high-risk scenes, thereby reducing liability costs while maintaining the ability to perform those sequences.
Similarly, the WGA’s protections against training AI on writers’ material preserved the union’s right to argue that such training violated copyright law or labor agreements, but did not contain explicit contractual language explicitly prohibiting the practice. The Copyright Office, as of 2026, was still conducting studies on whether and how copyright principles applied to AI training; federal courts continued deliberating on whether “fair use” doctrine protected companies that trained AI systems on copyrighted works. These legal ambiguities created space for studios to interpret agreements narrowly.
Real-world enforcement difficulties emerged. SAG-AFTRA members reported in 2024 and 2025 that they were being pressured to sign individual contracts consenting to digital replica creation—a practice that, while technically compliant with union rules, shifted the negotiating terrain from collective to individual, as actors faced a choice between refusing consent and losing employment opportunities. No widespread formal grievances were filed against major studios for breach of the digital replica protections, according to available accounts, suggesting either that studios were technically complying with the agreements or that enforcement mechanisms were proving difficult to activate. The 2024-2025 SAG-AFTRA video game strike, which produced a separate agreement extending likeness protections to interactive media, indicated that initial 2023 protections had not resolved underlying tensions, merely established a baseline from which future negotiations would proceed.
What The Divergence Indicates
The apparent contradiction between Hollywood’s 2023 anti-AI labor position and its 2025-2026 AI hiring acceleration is not ultimately inexplicable. It reflects, instead, the operation of competing pressures on large media conglomerates: the need to secure labor peace through contractual agreements that address worker concerns; the need to respond to structural economic pressures by reducing costs and building competitive capabilities; and the need to defend intellectual property against external threats from AI companies.
The 2023 labor agreements did constrain studio behavior at the margins. They did establish rights for performers regarding digital likeness and created procedural requirements around AI use. But they did not reverse the fundamental economic pressures that incentivize studios to adopt AI technologies. The agreements accomplished their primary objective—preserving negotiating power for creative workers and establishing precedent—without resolving the underlying dynamic: as production budgets contracted and legacy revenue sources declined, studios pursued efficiency through automation, and the displaced workforce became available for the precise kind of expertise that training AI systems required.
The hiring data from mid-2026—showing one in ten studio job postings connected to AI, alongside the emergence of platforms paying $40-100 per hour to displaced Hollywood creatives for AI training—suggests that the industry is transitioning toward a bifurcated labor market. On one side, a reduced cohort of higher-paid specialists working in AI development and IP defense. On the other, a much larger pool of gig-work AI trainers, each providing small increments of expertise to systems that collectively diminish the market for traditional employment. Neither outcome was explicitly mandated by the 2023 labor agreements; both have emerged from the economic logic that those agreements were negotiated against but did not fundamentally alter.
For the million people in Southern California whose livelihoods depend on film and television, this transition represents less a violation of the 2023 agreements than a reality that outpaced them—the emergence of a structural employment crisis that labor protections could constrain but not reverse. The question facing the industry and its workers going forward is whether that framework remains adequate, or whether it requires more fundamental alteration of how production is organized, valued, and compensated in an AI-augmented era.
Enjoyed This Story? Subscribe to EntertainLens for more entertainment insights.
