San Francisco-based fintech startup Slash has highlighted the financial vulnerabilities of generative AI adoption after an employee accidentally expended approximately $80,000 in development credits building a video game. Nicolas Brilliante, the firm’s head of strategic verticals, utilized the company’s AI coding infrastructure to construct a browser-based shooter titled Brainrot shooter. The project was developed during a period when management had actively encouraged staff to experiment with rapid, informal software creation workflows. Internal usage dashboards revealed that Brilliante’s initiative consumed $81,267 in tokens within a single week, a figure the company publicly confirmed and humorously suggested might be reclassified as a marketing expense. The incident rapidly circulated through technology and cryptocurrency networks, drawing attention to the cost implications of token-driven AI workflows. Brilliante acknowledged the expenditure was an unforeseen miscalculation, stating that he underestimated the computational resources required to complete the build. Despite the financial overrun, the episode reflects a broader corporate shift toward AI budget governance. As engineering teams increasingly integrate generative models into their pipelines, major technology firms have begun implementing strict spending controls to mitigate resource misallocation. Companies including Uber, Coinbase, and Walmart have recently established enterprise-wide caps on AI usage, citing concerns that unmonitored experimentation frequently yields negligible productivity gains while generating unpredictable costs. The Slash case now serves as a prominent reference point in ongoing industry discussions, prompting technology leaders to reassess how they balance creative AI utilization with financial oversight and operational accountability.