The State of AI 2026

Executive summary preview

The strongest pattern is a shift from AI tools that save staff time towards systems embedded in paid services, regulated workflows and daily operations, with stronger requirements for proof, control and accountability.

The most consequential shift is towards AI that can be purchased, integrated, governed and held accountable for a defined result.

The practical test is whether procurement conditions turn into signed contracts, pilots become renewals and operating outcomes are measured. Watch the economics of inference and the distinction between forecast regulation and obligations that have actually become binding.

Key findings

• AI use is broad, but material financial value remains concentrated in a small share of organisations.
• The commercial frontier is execution: action models, inference infrastructure and workflow automation rather than answers alone.
• Buyers increasingly ask for auditable performance, data residency, vendor neutrality and defined commercial terms.
• Permissions, audit trails and lifecycle governance are becoming part of the product.
• Capital is large but selective: AI took 64% of Q3 2026 venture funding, while cloud, compute and governance themes slowed in the latest 30 days.

Research figures

88%

of organisations use AI in at least one function

6%

get 5%+ of EBIT from AI

$725bn

Big Tech capex plans for 2026

What to watch in AI markets

1. Procurement clauses becoming signed contracts
2. Proofs of concept converting into renewals
3. Measured operating outcomes, not time-saved estimates

2 more items fully covered in the report.

Get the free summary