The headline “filmmakers can have a little AI, as a treat” that many media outlets quickly opted for after the Golden Globes announced their new AI qualifying guidelines on May 7 was more accurate than the actual wording. The policy is really well-considered. On closer inspection, it’s also based on a distinction that is so arbitrary that it’s difficult to see how anyone could consistently enforce it. According to the guidelines, AI does not automatically disqualify a movie, TV series, or podcast “provided that human creative direction, artistic judgment, and authorship remain primary throughout the production process.” Before you start wondering what any of those phrases actually imply in a cutting room, the sentence seems firm.
The policy’s main points are sensible enough to warrant serious consideration rather than ridicule. A performance must be primarily based on the work of a credited human performer in order to qualify for acting categories. AI can improve, such as de-aging, correcting a shot, or cleaning up a spoken line, but only with the performer’s consent. It cannot substitute an actor or utilize a digital likeness without permission. Performances that are entirely artificial are no longer available. The work remains eligible for the craft categories—writing, directing, composition, and others—as long as the “core creative contributions” are mostly attributed to humans, with AI playing a supporting role. This isn’t crazy at all. The rights that the WGA and SAG-AFTRA campaigned for during the 2023 strikes are mostly reflected in it.
The gap between “enhancing” and “creating,” which is where the entire policy resides and where it lacks a solid foundation, is the issue. Think about de-aging, which is specifically allowed under the regulations. In 2019, Robert De Niro’s de-aging in The Irishman was a digitally altered human performance. Alright. However, in 2026, generative video tools will be able to create full response shots, micro-movements, and expressions that an actor would never have made. When does “generating a performance the performer never gave” turn into “enhancing the credited performer’s work”? At this point, the policy makes gestures. It doesn’t define it, and perhaps it can’t.
The most important aspect to pay attention to is the disclosure requirement, which is the section with teeth. Any generative AI utilized in the production must now be disclosed in every submission. That one sentence establishes a paper trail that was previously nonexistent in the sector. It will be necessary for studios to monitor who authorized, where, and for what purposes AI was utilized. Once that documentation is standardized, it can be utilized in future eligibility disputes, labor negotiations, and legal proceedings. The disclosure system may prove to be the most significant action taken by the Globes, practically by accident, even if the substantive rules prove to be unworkable. Records tend to outlive the regulations that gave rise to them.
However, the optimism ends with the enforcement model. Submissions will be evaluated by the Golden Globes Eligibility Committee “based on the extent to which creative direction, artistic decision-making, and execution originate from credited individuals.” That is a committee’s subjective assessment that was primarily based on studio self-reporting. Measurable thresholds do not exist. There is no percentage of AI use that results in disqualification.
There is no technological detection mechanism described in the document. Although the committee has the authority to demand further information and penalize studios that refuse to comply, the core design is predicated on the assumption that those submitting films will truthfully explain the production process. That’s a lot of faith to give in a field where awards campaigns frequently cost millions and there is a strong incentive to distort the facts.
It’s difficult to avoid drawing comparisons between this and the Academy’s ongoing discussion over tighter regulations, which are getting closer to requiring screenplays to be written by people and acting to be performed by people. Instead, the Globes opted for flexibility, a decision that can be justified.
Being the awards body that supports studio innovation is a strategic position, not simply a philosophical one, as the HFPA’s replacement organization has been working for years to restore credibility following the scandals that almost killed the program. Studios are investing heavily on generative tools. If an awards presentation completely prohibits them, it runs the risk of becoming irrelevant. If an awards ceremony ignores them, its credibility is in jeopardy. The Globes divided the difference and hoped that no one would examine the seam too closely.