Skip to content
AI ConnectPowered by VELENTIS
AI-generated2 min

According to Deloitte Study: Majority of Enterprises Fail to Track AI Unit Costs

A global Deloitte survey of 3,500 executives shows that 56 percent fail to track AI workflow unit costs, while 40 percent permit autonomous actions despite liability ambiguities.

This article was AI-generated and published automatically. Context, labelling and all sources at the end of the article.

(KI-generiertes Symbolbild: Gemini / AI Connect)

As enterprise artificial intelligence transitions from early experimentation into daily business operations, structural governance gaps and cost blind spots are becoming increasingly visible. According to Deloitte's global survey "The State of AI in the Enterprise 2026", which polled 3,500 executives from director to C-suite levels across 25 countries and six core industries, organizations are deploying generative tools across many routine tasks. However, only 25 percent of enterprises have fundamentally transformed their core operating models. An overwhelming 81 percent of respondents merely report faster decisions and workflows, relying on technology to accelerate existing habits rather than redesigning them.

This narrow focus on speed without structural change is producing significant financial blind spots. While 71 percent of organizations now monitor their aggregate monthly expenditures for model inference and language model APIs, granular resource consumption remains largely unmeasured. Only 44 percent of enterprises track unit-level inference costs per specific workflow or business transaction. Without measuring these granular unit costs, finance departments lack the necessary benchmarks to determine whether specific AI use cases deliver positive economic returns or simply inflate overhead.

The financial visibility gap is further exacerbated by an acute lack of real-time budgetary controls. An accompanying survey of chief financial officers conducted alongside the main study revealed that just 7 percent of organizations possess automated capabilities to intervene in real time when inference budgets are breached. When runaway prompt loops or unexpected query volumes drive up API expenses, financial leaders typically discover the cost surge weeks later upon receiving aggregated cloud invoices. Proactive rate limiting and automated spending caps remain absent across the vast majority of enterprise infrastructures.

Despite these administrative shortcomings, monetization metrics show signs of gradual improvement across the corporate landscape. The share of enterprises attributing verifiable revenue growth directly to their AI deployments rose from 20 percent earlier in 2026 to 28 percent in the latest survey. This upward shift demonstrates that generative architectures can generate top-line value when effectively aligned with commercial workflows. At the same time, it widens the competitive gulf separating value-generating frontrunners from the 72 percent of companies that cannot yet prove direct commercial returns.

Concurrently, organizations are granting unprecedented operational authority to autonomous software systems without establishing clear legal guardrails. Fully 40 percent of surveyed enterprises now permit AI systems to execute operational actions and business decisions without requiring human sign-off. However, this shift toward autonomous agency has outpaced corporate risk management. In 23 percent of surveyed companies, complete internal ambiguity persists regarding who bears personal or legal liability when an autonomous agent triggers an erroneous action or damages business relationships.

The Deloitte findings signal a decisive pivot away from consequence-free AI exploration toward operational accountability. Corporate leaders must move past superficial velocity metrics and construct robust financial monitoring alongside enforceable governance policies. Organizations that fail to institutionalize workflow unit costing and define accountability for autonomous agent decisions risk accumulating unsustainable infrastructure expenses alongside unhedged legal vulnerabilities.

What this means for you

For IT directors and project leads, this study highlights that scaling AI tools without workflow-level unit cost tracking poses serious balance-sheet risks. Granting operational autonomy to agents without real-time API budget caps and clear legal responsibility leaves enterprises vulnerable to compounding technical and operational liabilities.

Evidence

Well sourced
83/100
  • According to Deloitte, 81 percent of executives report faster decisions through AI, yet only 25 percent have fundamentally transformed their operating models.

    verified
  • While 71 percent of companies track gross monthly AI spend, only 44 percent measure workflow unit costs, and only 7 percent of CFOs can counter budget overruns in real time.

    verified
  • The proportion of corporations reporting verifiable revenue gains directly from AI increased from 20 percent early in 2026 to 28 percent.

    verified
  • 40 percent of enterprises permit AI systems to execute operational decisions without human sign-off, while 23 percent report ambiguity over error liability.

    verified

The evidence score is computed, not hand-set: from confidence, the number of sources and the share of verified statements.

Source & transparency

As of: November 21, 2026

AI-generatedAI-generated: produced automatically from vetted sources with technical quality checks (source, quote and figure verification); no human sign-off of each item before publication

Sources
3
Verified statements
4 / 4
Evidence score
83Well sourced

Want to put this into practice?

We connect you with suitable AI providers from the DACH region, free of charge and without obligation.

What's next?