Enterprises slow AI rollouts as costs and oversight rise

Enterprise AI strategies are shifting from broad experimentation to tighter governance as costs, regulations and adoption gaps mount. The “use-AI-for-everything” mindset many companies adopted in 2025 and into 2026 has become harder to justify as AI providers move from flat-rate subscriptions toward consumption-based pricing and token use becomes a larger budget concern.

Gartner projected spending on AI models and platforms will increase 63% from last year, reaching $64 billion. Consultants said more advanced models are becoming more expensive to operate, with agentic models or agents requiring three to five times more tokens for a single query than earlier models. Enterprises are also contending with vendor fragmentation, inconsistent pricing and pressure to connect AI spending directly to revenue, customer experience or productivity gains.

Policy risk is rising as well. President Donald Trump signed an executive order in May establishing a voluntary review of frontier AI models for safety vulnerabilities, while the EU’s AI Act has brought core transparency rules, oversight powers, prohibitions on unacceptable uses and AI literacy obligations into effect for companies operating in Europe.

Workforce readiness remains another barrier. Forrester found only 16% of information workers had a high understanding of AI tools, while 43% were at risk of misunderstanding them, and 42% sometimes avoided AI because of ethics, privacy or business risk. Analysts said CIOs should audit token consumption, narrow access, consider lower-cost or domain-specific models and invest in hands-on training for nontechnical staff.