In brief:
AI cost management is becoming one of the next major FinOps challenges. Token-based pricing, model routing, caching, workload placement, and vendor consumption models are creating a new layer of complexity for enterprise technology leaders. The standards are still emerging, but the spend is already real, which means organizations need to start building visibility, governance, and commercial discipline before today’s experiments become tomorrow’s uncontrolled run rates.
Walking out of FinOps X, one thing was clear: the most honest conversations happening in enterprise technology right now are about what nobody knows yet.
The AI cost problem is real, it’s accelerating, and the frameworks to manage it are still being invented. Generative AI and agentic AI are moving from pilot projects into products, workflows, and business processes. Every prompt, retrieval step, model call, generated output, evaluation, and agent loop carries a cost. At small scale, those costs can look harmless. At enterprise scale, they become a commercial, operational, and governance issue.
According to the 2026 Gartner® First Take: Tokenomics Foundation Signals a Turning Point in Taming Runaway AI Costs, “leading organizations are increasingly taking a more considered, posthype approach that requires economic AI sustainability.”*
The challenge is, no-one truly knows how to solve this yet. That is not a criticism. It is the point. The organizations that lean into the uncertainty now will be better positioned when standards, benchmarks, and buying models start to settle.
My key learnings from FinOps X 2026:
1. The metrics don’t exist yet
Every organization consuming AI at scale is measuring something. Almost none of them are measuring the right things with real confidence yet. The industry has not landed on what good looks like for token consumption, AI value, or efficiency across different models, vendors, workloads, and use cases.
The danger is familiar: once a metric gets widely adopted, it gets gamed. Then it stops reflecting anything real. Cost per token may be useful, but it will not tell the whole story. Cost per query may be better for some use cases. Cost per successful outcome may be better again. The right answer depends on what the AI system is actually trying to do.
This is not a reason to stop measuring. It is a reason to stay skeptical of your own dashboards and invest now in better frameworks before the wrong standards calcify. The organizations doing that work today will be ahead when shared definitions do emerge.
2. Model routing is more dangerous than it looks
Routing workloads to cheaper models sounds like straightforward cost optimization. It isn’t. Route to the wrong model and you may break your cache, increase latency, reduce quality, or trigger a rebuild that costs more than the model selection saved in the first place.
That is what makes AI cost management different from traditional cloud cost optimization. You cannot optimize token cost at one layer and assume the system improves. Model routing, prompt architecture, caching, abstraction, workload placement, quality thresholds, and business outcomes all interact. If you only optimize the unit price, the savings can backfire.
3. Vendor pricing opacity is a business model risk
Software vendors are shifting away from familiar seat-based models and toward consumption, credit, and usage-based pricing. In theory, that should align cost with value. In practice, many of these models make true cost harder to understand.
Credits deplete quickly. The math is hard to follow. Unit definitions vary. And the downstream effect is real. If your vendors change their token pricing, usage limits, or consumption rules, it may force you to rethink the economics of your own products, services, internal workflows, or AI-enabled customer experiences.
The window to push for contractual visibility is now, before these models become the default and leverage disappears. Enterprises should be asking for clearer consumption terms, better reporting, more transparent unit economics, and the ability to attribute AI usage back to teams, products, and business outcomes.
4. The industry is organizing around this problem
The launch of the Tokenomics Foundation is the clearest signal yet that AI cost management is being taken seriously at an institutional level. The goal is to create open standards, benchmarks, and best practices for the economics of AI infrastructure, working closely with the FinOps Foundation as token-based AI becomes a new form of variable technology spend.
This problem sits on both sides of the AI economy. Buyers need transparent, vendor-neutral standards for AI consumption. Suppliers need clearer ways to define, price, benchmark, and explain the economics of the infrastructure they are selling. Neither side benefits from a market where the bill arrives before anyone understands the model.
5. The open questions are the work
The most important thing about where this space stands is the list of questions that don’t have answers yet:
- How do we standardize metrics across wildly different token pricing models?
- How do we price uncertainty, retries, and failure in AI experiments?
- Who gets token budget, and how do you decide?
- Should humans be doing some of this work at all?
- When should a workload use a frontier model, a smaller model, a cached response, a rules-based workflow, or no AI at all?
- How do we connect AI consumption to business value rather than just usage?
These are not hypothetical questions. They are the actual decisions organizations are making right now without a mature framework to guide them.
What enterprise leaders should do next
The worst move is to wait for the market to mature before taking action. AI spend is already moving faster than the operating models around it. Leaders do not need perfect standards to start building better discipline.
- Start with visibility. Know where AI is being used, which models are being called, who owns the workload, what business outcome the usage supports, and how costs are being allocated. Then bring FinOps, procurement, engineering, finance, legal, risk, and business stakeholders into the same conversation. This cannot sit with one team.
- From there, push for transparency. Ask vendors how usage is measured, how pricing can change, what reporting is available, and whether consumption can be mapped to internal teams, products, or cost centers.
- Build internal guardrails before AI adoption becomes too distributed to govern. And do not separate cost from quality, performance, security, or value. In AI, the cheapest option can quickly become the most expensive one.
The FinOps community is at the same place it was with cloud cost in 2012. The spend is real, the pain is real, and the discipline is being built in real time. The practitioners investing in this now are the ones who will define what it looks like in five years.
NEXT STEPS: Speak with an SHI expert and let’s work together to solve this emerging challenge.
Want to learn more on this topic? Read our blog on FinOps for AI – how to stop chasing tokens and start measuring outcomes
*Gartner, First Take: Tokenomics Foundation Signals a Turning Point in Taming Runaway AI Costs, 5 June 2026, GARTNER is a trademark of Gartner, Inc. and/or its affiliates.
PakarPBN
A Private Blog Network (PBN) is a collection of websites that are controlled by a single individual or organization and used primarily to build backlinks to a “money site” in order to influence its ranking in search engines such as Google. The core idea behind a PBN is based on the importance of backlinks in Google’s ranking algorithm. Since Google views backlinks as signals of authority and trust, some website owners attempt to artificially create these signals through a controlled network of sites.
In a typical PBN setup, the owner acquires expired or aged domains that already have existing authority, backlinks, and history. These domains are rebuilt with new content and hosted separately, often using different IP addresses, hosting providers, themes, and ownership details to make them appear unrelated. Within the content published on these sites, links are strategically placed that point to the main website the owner wants to rank higher. By doing this, the owner attempts to pass link equity (also known as “link juice”) from the PBN sites to the target website.
The purpose of a PBN is to give the impression that the target website is naturally earning links from multiple independent sources. If done effectively, this can temporarily improve keyword rankings, increase organic visibility, and drive more traffic from search results.
