Tokenomics emerges as companies struggle to prove AI returns

Companies were on track to spend as much as $680 billion on AI software and models in 2026, Reuters reported in June 2025, a bill big enough to force finance chiefs to ask what the money was actually buying. That pressure has given rise to “tokenomics,” a shorthand for turning AI usage into measurable business returns instead of another line item in the technology budget.

McKinsey’s State of AI survey showed how quickly adoption had spread. In early 2024, 65% of respondents said their organizations were regularly using generative AI, nearly double the share from ten months earlier. By March 2025, McKinsey said companies were starting to build the structures and processes needed to capture value from gen AI, with large companies leading the way and reporting early cost reduction and revenue growth benefits.

Finance leaders have been more cautious about calling those benefits durable. Deloitte’s March 18, 2024 CFO Signals survey found more than half of the CFOs surveyed saw potential for higher productivity from genAI, even as they worried about skills gaps inside their organizations. Salesforce’s 2025 CFO research went further, saying chief financial officers were redefining ROI as AI agents became more strategic in their companies. Bain & Company has made a similar argument, saying the CFOs who funded the AI revolution are now joining it.

Some vendors are already putting hard numbers on the table. Thomson Reuters has highlighted a professional-services customer story in which Justly Prudent said it achieved 5 times the ROI with CoCounsel and increased litigation capacity by 100%. Those are the kinds of figures AI sellers now lean on as executives demand evidence that pilot projects, legal tools and enterprise assistants are producing more than impressive demos.

The problem is that broad proof remains elusive. A World Economic Forum article citing a recent MIT study said 95% of enterprise AI initiatives were not producing measurable returns, a stark warning that many deployments are still failing to move from experimentation to sustained business value. That gap between soaring spending and uneven payback is what tokenomics is meant to close, and what corporate boards will keep pressing on as the AI boom moves from hype to accounting.