Just when IT and finance teams thought they had cloud costs figured out, AI came in and introduced a new layer of complexity. AI budgets are growing faster than IT budgets , and GenAI spending is expected to jump by 76% in 2025. This rapid growth is fueling increased complexity and cloud waste, a problem so significant that the FinOps Foundation has now added AI cost tracking to its framework.
The costs of AI can spiral out of control for several reasons. AI workloads run on expensive, GPU-intensive infrastructure that can cost up to $50 an hour. Many enterprise teams are also using AI tools that charge per API request, with hundreds of thousands of interactions adding up fast. On top of that, "shadow AI" services—unmanaged subscriptions purchased by individual teams—often go undetected.
To keep AI costs in check, organizations need a new strategy. The blog post from Tangoe highlights several best practices, including:
- Detecting and tagging AI workloads from the start.
- Tracking the usage of API-based models and attributing costs to specific teams.
- Breaking down and monitoring GPU costs separately from regular cloud spend.
The answer to this challenge lies in a FinOps Certified platform that can provide deep visibility and targeted controls. Tangoe One Cloud is purpose-built for
AI cost optimization and financial oversight. It features an AI Cost Visibility Dashboard to help teams spot usage spikes and provides actionable insights to detect underutilized resources, with the potential to save up to 40% on cloud costs.
See how much you could be saving with a demo of Tangoe One Cloud.
Tangoe is a proud Senior Engagement Partner of the CxO Institute event in Oxford
They are committed to helping enterprise leaders drive innovation with financial accountability. To see how Tangoe One Cloud empowers IT and finance leaders to tackle AI costs head-on, schedule a demo today.