Technology is scaling faster than our ability to govern it. Software licences, SaaS subscriptions, cloud consumption and now AI tokens are all growing at once, bought by different teams, billed in different formats and owned by nobody in particular. That is not a tooling problem. It is a capability crisis, and most companies haven't seen it yet.
Three disciplines, one problem
ITAM, FinOps and AI governance grew up as separate practices. In reality they answer the same question: what do we have, what does it cost, who uses it, and what value does it create?
- IT Asset Management (ITAM) keeps control of software and hardware entitlements, licence compliance and renewals.
- FinOps brings financial accountability to variable cloud spend, so engineering, finance and business make cost-aware decisions together.
- AI governance is the newest layer: which models are used, where they run, what they cost per token, and whether they are used responsibly.
The lines between them are disappearing. An AI coding assistant is a SaaS licence (ITAM), consumes tokens billed on usage (FinOps) and processes company data through a model (AI governance). Managing it in three silos means three partial pictures and no single answer for the CFO.
What the data says: value, visibility and fragmentation
The new State of Tokenomics report from the Tokenomics Foundation (September 2026, 472 large companies across 11 industries, median revenue $1.8B) confirms the pattern. AI spend is growing fast, but the ability to govern it is not.
| Finding | Share of respondents |
|---|---|
| Cannot confidently prove AI business outcomes to the CFO | 3 in 4 |
| Name proving value / ROI as their biggest challenge | 43% |
| Name visibility and attribution of spend as a top challenge | 27% |
| Very confident in their AI spend visibility | 14% |
| No defined owner of AI economics | 12% |
| Use AI across 4 or more of 6 procurement and hosting channels | about half |
Three themes stand out:
- Value. Most organisations cannot connect AI investments to business outcomes beyond basic token counts.
- Visibility. Spend is spread across model providers, cloud platforms, embedded AI in tools, and private or rented hardware, with no unified view across vendors and business units.
- Fragmentation. No common metrics across providers make consistent governance and forecasting hard.
Tellingly, only 7% named price as a challenge, and only 4% asked providers for cheaper prices. What enterprises want is transparency, standards such as FOCUS, and attribution. That is a governance agenda, not a procurement one.
The cost of not governing
Without joined-up governance, technology spend becomes unpredictable, and unpredictability is what boards and CFOs punish.
- Budget overruns. Usage-based AI and cloud costs move month to month. Spending caps and token limits, the most common control today, stop the bleeding but say nothing about value.
- Margin pressure. 52% of companies have already changed, or are considering changing, their own pricing because of AI costs.
- Blind spots at scale. Confidence falls as companies grow. The bigger the estate, the less anyone sees the whole picture.
- Compliance and risk. Shadow AI, unlicensed software and unmanaged data flows create audit exposure that surfaces at the worst possible moment.
The good news is that governance pays off measurably. In the same report, companies with defined ownership of AI economics are 3.7x more likely to show value to the CFO. Those that route workloads to the right model are 4x more likely. Control is not the opposite of innovation; it is what lets innovation be funded again next year.
How SofClo helps: control, efficiency, predictability
SofClo brings ITAM, FinOps and AI governance together in one AI-powered platform, run as a Tech Value Program. Instead of three teams with three spreadsheets, you get one view of your technology estate and one language for its cost and value.
- Control: one source of truth. Software, SaaS, cloud and AI consumption in one place, attributed to the teams, projects and business units that use them. You know what you have, what it costs and who owns it.
- Efficiency: act on what the data shows. Identify unused licences, oversized cloud resources and AI workloads running on more expensive models than they need. Turn insight into recurring savings, not one-off clean-ups.
- Predictability: know what needs to be governed next. Trend and forecast spend across licences, cloud and tokens, so renewals, budgets and AI scale-ups are planned rather than discovered on the invoice.
- Value: a story the CFO believes. Connect technology spend to business outcomes, so AI and IT investments can be defended with facts.
SofClo is built for European organisations: AWS-native, ISO 27001 certified, and designed by people who have spent decades helping companies get control of, and value from, their technology estates.
Five first steps for technology and finance leaders
- Name an owner. Make one person or forum accountable for the economics of software, cloud and AI together.
- Map the estate. List every channel through which technology is bought and consumed, including AI embedded in tools your teams already use.
- Attribute spend. Tag cost to teams, products and outcomes, not just to vendors.
- Define value metrics. Agree up front what good looks like: tickets resolved, time saved, revenue enabled.
- Forecast, then govern. Use trends to decide where policy and controls are needed, before the invoice decides for you.
Let's talk in Amsterdam
The organisations that win the next five years won't be the ones with the most AI. They will be the ones that can see, steer and prove the value of everything they run.
We will be at SAMS Europe / ITAM Europe in Amsterdam on 28–29 October 2026. Come and talk to us about getting control of your ITAM, FinOps and AI estate, or book a conversation at www.sofclo.com.
Source: State of Tokenomics, Tokenomics Foundation, 23 September 2026 (Release Candidate 1.0), tokeneconomics.com.