Quilty wants to turn screenplay evaluation into a score, but its first public tests show why founders should be careful when they sell prediction as certainty.
The useful part of Quilty is not the claim that an AI system can spot a hit movie before cameras roll. The useful part is what the company reveals about the market. Studios, producers and financiers are still trying to reduce creative risk, and AI startups are rushing to offer them a number they can use in the room.
That is a familiar founder story. Take a messy, expensive decision. Wrap it in software. Make the output simple enough for a buyer to understand. Quilty, launched in March 2026 by Simon Horsman and Daniel Wood, is doing that for entertainment development. The platform analyzes scripts and produces a Quilty Score from 0 to 100 across story and craft, commercial viability, cultural resonance and production reality.
On paper, that sounds exactly like the kind of workflow AI should improve. Script coverage is slow. Development notes are inconsistent. Financing decisions often depend on relationships, instinct and timing. A tool that can organize comparable titles, production assumptions, talent suggestions and market positioning has obvious value. The problem begins when that decision-support layer starts sounding like a crystal ball.
As The Verge recently noted in its coverage of Quilty, the company is pitching its system as a way to predict whether a screenplay has real commercial potential before production begins. Quilty says it uses multiple AI models and a sentiment engine to analyze not only story mechanics, but also the way a project might land with audiences today.
That last part is where the promise gets harder to defend. Culture does not behave like a spreadsheet. A film can win because of timing, marketing, casting, memes, controversy, scarcity or simply because audiences are ready for something executives did not know they wanted. The script matters, but it is one part of a chain that includes financing, distribution, campaign strategy, release windows and luck.
TheWrap ran a practical test in April using scripts for Barbie, Sinners, Christy and Die Hard. Its results were not flattering. Quilty reportedly gave Christy the highest score among the four, ahead of Barbie and Sinners. That matters because Christy, the Sydney Sweeney boxing biopic, grossed only a few million dollars, while Barbie became a $1.4 billion global phenomenon and Sinners crossed roughly $369 million worldwide.
That does not mean Quilty is useless. It means the score needs to be treated as a prompt for discussion, not a substitute for judgment. A model can surface risk. It can compare budget assumptions. It can flag a difficult production path. What it cannot do reliably is explain why a film becomes the one people talk about that summer.
For founders, the category is real but narrow
There is a fundable idea here, but it may not be the one that gets the loudest headline. The market for AI in entertainment is moving beyond image and video generation into planning tools, workflow systems and financial intelligence. Producers do not only need scripts reviewed. They need budgets modeled, tax incentives compared, schedules tested, buyers mapped and investor updates cleaned up.
That is where Quilty may have a better business than its most ambitious language suggests. A $49.99 in-depth analysis gives writers and smaller producers access to something that used to look like studio infrastructure. A broader platform for packaging, market forecasting and production planning could become useful if it saves time and brings more discipline to early decisions.
But the risk for a startup like this is absorption. Script analysis by itself looks more like a feature than a company. Larger entertainment-tech platforms, production management systems and studio data vendors can add AI coverage, comparable-title analysis and forecasting modules without needing to make prediction the whole brand. If the product becomes a workflow hub, it has more room to stand. If it remains a score, it will be easy to copy and easy to dismiss.
This is the lesson for entrepreneurs outside Hollywood as well. AI products that support high-stakes capital allocation need validation that buyers can trust. That means showing how predictions are tested, what inputs are used, where the model fails and how often the tool changes a decision for the better. A clever interface is not enough when the customer is being asked to risk real money.
Studios will not stop using data because artists dislike it, and they will not stop relying on taste because software produces a cleaner report. The likely future is more mixed. Executives will use AI to process more material, compare more options and document more assumptions, while still relying on people to decide whether a project has life beyond its metrics.
That is the practical takeaway. The opportunity is not to replace judgment. It is to make judgment better informed. Quilty may still become a useful company if it leans into that role, but the market will be far less forgiving of any founder who confuses a polished score with proof.
Also read: Europe's Qwant switch gives local search startups a real opening • Shield AI's drone incidents turn defense-tech speed into an investor test • Microsoft's Scout leak has turned AI stickiness into a boardroom risk
