Startup employee accidentally spent Rs 75 lakh on AI coding in 7 days while making a game

A startup employee accidentally ran up a massive AI coding bill while building a simple shooter game, turning a quick experiment into a costly surprise. The incident shows how unchecked AI usage in coding tools can quickly spiral into huge expenses for companies.

In Short

• Employee at fintech startup Slash spent Rs 75 lakh on AI tools in just 7 days
• The money was used while building a basic meme-style shooter game called “Brainrot Shooter"
• The case adds to growing concerns over rising AI coding costs and companies now tightening usage limits

A startup employee's experiment with AI-powered coding has turned into a costly lesson after he accidentally spent nearly Rs 75 lakh worth of AI credits in just one week while building a simple online game. The incident involves San Francisco-based fintech startup Slash, which recently revealed that one of its employees ran up an AI bill of more than $80,000 while creating a game called "Brainrot Shooter." The company shared the story in a post on X, saying it had encouraged employees to spend more time "vibe coding," a term often used for building software quickly with the help of AI tools. However, things escalated much faster than expected.

"We encouraged the company last week to start vibe coding more but @nickbruhman burned $80k in credits on the Slash card for a brainrot shooter," the company wrote on X. It jokingly added, "Pls play it so we can write this off as a marketing expense."

The game itself is fairly basic. Set in a block-style world similar to Minecraft, it asks players to shoot internet meme-inspired characters with names such as "skibidi toilet" and "tung tung tung sahur." The employee behind the project, Nicolas Brilliante, later shared what appeared to be a screenshot of his AI usage dashboard. According to the image, he had consumed $81,267 worth of AI services while building the game.

"This was a genuine accident, I underestimated my own ability," Brilliante wrote on X.

As the story gained attention online, prediction market platform Polymarket described the situation as an example of how AI coding costs can spiral when usage is left unchecked. Reacting to the growing discussion, Brilliante himself wrote, "This is actually insane, am I going to become a case study for how AI spend can get out of control."

Companies tighten AI spending as costs rise

The incident comes at a time when many companies are closely examining how much money they are spending on AI tools. While AI coding assistants promise faster software development, some businesses are finding that the costs can rise quickly, especially when employees use powerful models extensively.

In fact, this is not the first example of AI spending getting out of hand. Recent reporting cited an unnamed company that reportedly spent around $500 million in a single month on Anthropic's Claude AI services after failing to set usage limits for employees. The case has become one of the most talked-about examples of how unrestricted AI access can create massive bills.

As AI adoption grows, several large companies have started introducing spending controls. Reports indicate that firms including Uber, Coinbase and Walmart have implemented limits on employee AI usage.

Bloomberg reported that Uber recently introduced monthly caps on AI coding tools after reportedly exhausting its annual AI budget within the first few months of the year. Walmart has also tightened controls on AI usage to reduce unnecessary coding experiments and repetitive tasks that do not provide meaningful business value.

The concern for businesses is whether the rapidly growing AI bills are producing enough productivity gains to justify the spending. While many technology leaders believe AI will eventually improve efficiency, companies are increasingly looking for evidence that the investment is translating into real business results.

- Ends