After almost five months, the Writers Guild of America (WGA) and Hollywood studios reached an agreement to end the writers’ strike. The contract sets terms for returning to work, including rules for how artificial intelligence can be used in writers’ rooms.
The contract draws a line around scriptwriting
Under the agreement, AI cannot write or rewrite scripts. AI-generated writing also cannot count as source material, a provision intended to protect writers from losing credit because an AI tool contributed to a work.
These limits address a central concern during the strike: that studios might use fast-developing generative AI tools to reduce the work available to union members or avoid paying them. The article notes that text-based tools such as ChatGPT were considered creatively limited, but writers were concerned about how companies might deploy them.
Comedy writer Adam Conover described that concern at the start of the strike: “I’m not worried about the technology,” he told TechCrunch. “I’m worried about the companies using technology, that is not in fact very good, to undermine our working conditions.”
Writers retain a choice, and studios have duties
The agreement distinguishes between an individual writer choosing to use AI and a company requiring its use. Writers may use AI tools if they want, but a company cannot mandate that they use particular tools while working on a production.
Studios also have to tell writers when they are given AI-generated material to incorporate into a work. Together, these provisions clarify who makes the decision to use a tool and establish that writers should know when AI-generated content is part of the material they receive.
The contract summary also addresses the use of writers’ work to train AI. As the WGA summary states, “The WGA reserves the right to assert that exploitation of writers’ material to train AI is prohibited by [the contract] or other law.”
A union agreement amid legal uncertainty
The legal relationship between large language models and copyrighted material remains murky, according to the source article. Against that uncertainty, the WGA agreement states that union members do not consent to their work being used to train studios’ AI systems.
A negotiated contract can establish obligations for the parties covered by it even while broader legal questions remain unresolved. In this case, the terms set out boundaries for script generation, writing credit, required disclosure and the use of members’ material for AI training.
The agreement also sits within a wider debate about how creative work is treated when companies adopt AI. The writers’ concerns were not limited to the capabilities of the tools; they also focused on the possibility that studios could use them in ways that affect pay and working conditions.
Other performers are still negotiating AI protections
At the time of the article, the actors union SAG-AFTRA remained on strike. Its members had also voted overwhelmingly to authorize a strike against the video game industry, where the union represents stunt, motion capture and voice actors.
SAG-AFTRA had raised concerns that performers could be scanned or have their voices used to train AI soon after arriving for work. Its website warned: “For many performers, their first job may be their last, as companies become increasingly eager to scan our members or train AI with their voices as soon as they show up for work.”
The terms for SAG-AFTRA members had yet to be determined. The WGA agreement, however, offered an example of how a union contract can set limits on AI use in a creative profession. Its provisions make the question concrete: who can require a tool, how AI-generated material is handled, and what protections apply to workers’ contributions.