Will AI scripts predict box office hits or flops accurately?

Hollywood has long relied on gut instinct, producer hunches, and studio track records to greenlight projects. But a London-based AI startup called Quilty is betting its machine learning model can tilt the odds in favor of creatives and investors by predicting a film’s box office performance—or even Oscar potential—from nothing more than a screenplay.

When Quilty emerged in trade publications earlier this year, it positioned its tool as a groundbreaking solution for script analysis. The company argued that by ingesting vast datasets on past films—box office numbers, critical reception, genre trends, and even script formatting—its AI could quantify a script’s market viability before a single dollar was spent on production. Yet when filmmakers and industry analysts finally tested the platform, skepticism quickly set in.

How Quilty’s AI Script Analyzer Works

Quilty’s approach centers on a predictive model trained on thousands of released films. The system evaluates script elements such as pacing, dialogue structure, character arcs, and thematic consistency against historical data. It assigns scores for "market fit," "audience engagement potential," and even "award likelihood." According to co-founder Simon Horsman, the tool is designed to surface patterns that human readers might overlook—patterns that could correlate with financial success or critical acclaim.

The company frames its AI as a democratizing force, promising to level the playing field for independent filmmakers and emerging writers who lack access to studio networks. "We want to give creatives the same data-driven insights that big studios have always used," Horsman said in an interview. By removing some of the guesswork, Quilty claims, the industry could reduce costly misfires and greenlight projects with stronger commercial prospects.

Real-World Tests Reveal Mixed Results

But real-world testing has raised questions about the tool’s reliability. In one widely discussed case, Quilty’s AI scored the script for Christy—a low-budget boxing drama starring rising talent—as significantly more promising than the script for Sinners, a gritty crime thriller that later became an Oscar-nominated smash hit. The prediction proved wildly off-base: Christy underperformed at the box office, while Sinners earned critical praise and major nominations.

Critics point out that box office performance is influenced by factors beyond script quality—casting choices, marketing budgets, release timing, and cultural trends all play major roles. An AI trained on past data may struggle to account for black swan events or sudden shifts in audience tastes. Some industry veterans dismiss the tool outright. "No algorithm can predict whether a film will resonate emotionally with audiences," said one longtime producer who tested the platform anonymously.

The Broader Debate Over AI in Creative Industries

Quilty’s tool is part of a growing wave of AI applications in entertainment, from scriptwriting assistants to deepfake voice actors. Studios are experimenting with generative AI to draft treatments, while platforms like Netflix have explored AI-driven content recommendations. Yet skepticism persists about whether AI can truly understand—or enhance—human creativity.

For its part, Quilty acknowledges the limitations. The company says it’s refining its model based on user feedback and expanding its dataset to include more international films and streaming successes. Horsman emphasized that the tool is meant to complement, not replace, human judgment. "We’re not claiming AI can make a film a hit," he noted. "But if it helps creators avoid obvious pitfalls, that’s value enough."

As AI tools like Quilty’s continue to evolve, the entertainment industry will need to strike a balance between data-driven decision-making and artistic intuition. For now, the question remains: Can an algorithm really tell if your script will make a hit film? The evidence so far suggests it’s early days—and the human touch is still irreplaceable.