For one week inside a San Francisco fintech startup, a side project built on internet jokes quietly racked up a bill that could buy a luxury car. The developer, Nicolas Brillante, head of strategic verticals at Slash, spent a full day using Anthropic’s coding assistant Claude to build a game he called “Brainrot Shooter.” The game drew on viral memes including “Tung Tung Tung Sahur” and “Skibidi Toilet.” By week’s end, the company’s AI bill hit $81,267. Brillante posted a screenshot of his usage dashboard on X.
He called it an honest mistake. “This was a genuine accident, I underestimated my own ability,” he wrote.
In a reply to a Polymarket post on X, he added: “This is actually insane, am I going to become a case study for how AI spend can get out of control.” The project was part of a company push to encourage greater use of AI coding tools to accelerate software development. That push, meant to boost productivity, instead produced a costly lesson. Each time Brillante asked Claude to analyze and then update large blocks of code, the AI token count climbed.
Tokens are the units of text AI models process and charge for. The meter ran without any obvious cap.
The incident has ignited an online discussion about the hidden costs of AI-powered development. It also raises a more basic question: when a company tells engineers to use AI tools, who watches the tab? Slash is a fintech firm.
The employee building a meme game on company time — and on the company’s AI account — suggests a culture where experimentation was encouraged but guardrails were missing. Artificial intelligence is transforming how developers write software. Faster coding and greater productivity come with a price tag.
That price tag, in this case, was $81,267 for one week of one employee’s work. What began as a harmless experiment in productivity turned into a costly lesson for the firm.
Sources familiar with the matter note that the episode underscores the need for businesses to enforce stronger spending controls and limits on employee use of advanced coding tools. That is a polite way of saying: someone should have set a budget. The $81,000 figure is roughly the cost of a new Tesla Model S.
It is more than the median U.S. household income in a year. It is the kind of number that makes a finance team wince.
And it came from a single developer typing prompts into a chatbot. Brillante’s posts on X have drawn attention from other developers and tech observers. Some have shared their own stories of surprise AI bills.
Others have pointed out that the incident is not just about one person’s mistake — it is about how companies adopt powerful tools without fully understanding their costs. The record shows that the episode is a case study in what happens when AI spend gets out of control. The broader context matters.
AI coding assistants like Claude are marketed as productivity multipliers. They can write code, debug it, refactor it, and explain it.
They can save developers hours. But they charge per token, and complex tasks — like analyzing and updating large blocks of code — burn tokens fast. A developer who treats an AI assistant like a free intern will eventually get a bill.
The record shows that what happened at Slash is not unique. It is a preview of a problem that will hit more companies as AI tools become standard.
The hidden cost of AI-powered development is not just the subscription fee. It is the runaway usage that no one tracks until the invoice arrives.
{“@context”: “https://schema.org”, “@type”: “NewsArticle”, “headline”: “Developer’s Meme Game Built with AI Costs Startup Over $81,000”, “description”: “For one week inside a San Francisco fintech startup, a side project built on internet jokes quietly racked up a bill that could buy a…”, “datePublished”: “2026-06-26T12:36:00+00:00T09:30:00+00:00”, “author”: {“@type”: “Person”, “name”: “Merry Gel Sigui”}, “publisher”: {“@type”: “Organization”, “name”: “InfoPulse Today”, “url”: “https://newsfeed.infopulsetoday.com/”}}
