---
id: "lib-2026-01-00-state-of-ai-in-the-enterprise-2026"
type: "library"
title: "State of AI in the Enterprise 2026"
lang: "EN"
org: "Deloitte"
source_type: "Report"
date: "2026-01-01"
week: "2026-W01"
url: "https://nosov.ai/library/lib-2026-01-00-state-of-ai-in-the-enterprise-2026/"
original_url: "https://www.deloitte.com/content/dam/assets-zone2/uk/en/docs/about/2026/state-of-ai-2026-report.pdf"
tier: "B"
tier_name: "Useful, limited"
tier_reason: "Deloitte AI Institute's annual survey of 3,235 director-to-C-suite leaders in 24 countries; large and widely cited, but every figure is self-reported by the leaders running the AI programmes."
summary_basis: "full-text"
original_minutes: "35"
tags: ["enterprise ai", "talent", "workflow"]
---

# State of AI in the Enterprise 2026

Tier B (Useful, limited): Deloitte AI Institute's annual survey of 3,235 director-to-C-suite leaders in 24 countries; large and widely cited, but every figure is self-reported by the leaders running the AI programmes.

Summary based on the full text.

## In short

Deloitte's January 2026 State of AI in the Enterprise report, "The untapped edge". Worker access to sanctioned AI tools has widened, and most firms report efficiency gains, but few have redesigned work, revenue gains remain rare, and agent deployment is outrunning governance. It adds sections on sovereign AI and physical AI.

## Key numbers

- 84%: companies that have not redesigned jobs or the nature of work around AI
- 74%: companies planning to deploy agentic AI within two years
- 21%: companies reporting a mature governance model for autonomous agents
- 20%: organisations already growing revenue through AI, against 74% that hope to
- 25%: respondents whose organisation has moved 40% or more of AI experiments into production

## What it reports

- Worker access to sanctioned AI tools rose from under 40% to around 60% in a year. Among staff with access, fewer than 60% use it daily, unchanged from last year.
- 25% have moved 40% or more of AI experiments into production. 54% expect to reach that level in three to six months.
- 34% use AI to deeply transform the business, 30% redesign key processes and 37% use it at surface level.
- 66% report efficiency and productivity gains today. Only 20% report revenue growth, against 74% hoping for it.
- 84% have not redesigned jobs around AI. Worker skills are named the biggest barrier, yet fewer than half are changing talent strategies.
- 74% plan agentic AI within two years. 21% have a mature model for agent governance.
- 77% say the location of AI development matters when choosing technology. 58% already use physical AI to some extent.

## How they got it

Online survey of 3,235 director to C-suite respondents across six industries and 24 countries, fielded August to September 2025, plus 15 executive interviews. Methodology on page 40.

## Limits

- Respondents are directly involved in their companies' AI initiatives, so optimism bias is likely.
- No outcome data: productivity and transformation levels are self-assessed.
- The 25% to 54% production jump is an expectation over three to six months, not an observed result.

## Why it matters for companies

The gap between 60% access and fewer than 60% daily use among those with access is the number to check internally: licences are not usage. The 84% figure says most firms are training people on tools while leaving roles and workflows unchanged, which caps the gains at surface efficiency. Anyone planning agents should read the 74% versus 21% split as a warning to put ownership, permissions and audit in place before scaling.

Original: https://www.deloitte.com/content/dam/assets-zone2/uk/en/docs/about/2026/state-of-ai-2026-report.pdf