Cursor, Claude Code, and similar AI coding tools have surged, capturing 55% of all departmental AI spending, which amounts to $4 billion out of a total $7.3 billion in 2025. This rapid growth, from $550 million in 2024 to $4 billion in just one year, highlights a significant shift toward coding solutions in the AI landscape. The demand for these tools is driven by the measurable productivity gains they offer. Developers are increasingly adopting AI tools, with 50% using them daily, and in top organizations, that figure rises to 65%. The article breaks down the remainder of the $7.3 billion in AI spending. IT operations account for 10% ($730 million), focusing on automating tasks like incident response and infrastructure management. Marketing and customer support each make up 9% ($660 million and $630 million, respectively), with applications in content generation and ticket resolution. Design and HR follow, with smaller budgets of 7% ($510 million) and 5% ($365 million). While coding is leading the charge, other sectors are expected to grow, albeit at a slower pace. The turning point for coding tools came with Anthropic’s Claude Sonnet 3.5 in mid-2024, which made AI coding economically viable for enterprises. The measurable nature of coding—such as lines of code shipped and pull requests merged—makes it easier to justify investment. In contrast, ROI in areas like marketing or HR is harder to quantify, making it more challenging for those categories to achieve similar growth. Overall, coding tools have established themselves as the first major success in the generative AI domain, setting a precedent for other fields to follow.