A new report warns of a lack of financial visibility in artificial intelligence projects, despite strong growth in budgets and expectations.
Investment in artificial intelligence continues to accelerate, but its return remains uncertain. This is revealed in the first comprehensive report on cost efficiency in AI deployments, published by CloudZero, a company specializing in proactive cost optimization in the cloud. The study, titled The State of AI Costs, is based on surveys of over 500 software industry professionals in leadership positions and reflects a clear trend: AI budgets will increase by 36% in 2025, while only 51% of organizations say they can confidently evaluate the return on these investments.
The difficulty in measuring ROI is attributed, among other factors, to the complexity of isolating AI’s impact from other ongoing technological or strategic initiatives within companies.
More spending, but without clear control
The report confirms that companies are allocating more resources to artificial intelligence for various objectives: automating development processes, enhancing cybersecurity, complying with regulations, and driving innovation. However, the lack of cost governance could turn these projects into financial black holes without a clear return.
Key findings include:
- Increase in spending on AI: the average monthly budget for artificial intelligence will grow by 36% in 2025.
- Focus on automation and scalability: the most frequently used tools are aimed at cloud environments, but without effective financial traceability, their profitability is uncertain.
- Specialized hiring as a bottleneck: high salaries, a shortage of qualified talent, and lack of internal knowledge are the main challenges. The most in-demand skills are cloud computing and data engineering.
- Visibility improves confidence: companies using external cost optimization tools demonstrate a much greater ability to accurately calculate the ROI of their AI investments.
A maturation moment for enterprise AI
Erik Peterson, founder and CTO of CloudZero, summarized the current situation with a clear warning: “AI spending is following the same cycle of enthusiasm we’ve seen with other transformative technologies: many bets, little clarity. The difference this time is the speed and scale of spending.”
According to Peterson, the key will not be who spends the most on AI, but who can scale it efficiently. “Companies that manage to link AI use to real business outcomes are not only not slowing down spending, they are doubling down. Just as with the cloud, AI will soon be a structural part of any modern company. And when that happens, those who master the profitability metrics per unit of AI will win,” he concludes.
Financial governance, the pending task
The CloudZero report arrives at a time when many companies are adopting generative AI solutions, LLM platforms, and automated workflows without yet solidifying a clear cost control strategy. The lack of visibility and attribution of costs by project or business unit can pose a significant risk in terms of sustainable scalability.
CloudZero recommends an engineering- and finance-led approach to achieve a comprehensive view of operational spending in the cloud and transform AI investments into measurable competitive advantages.
The full report can be accessed at: cloudzero.com/state-of-ai-costs