Survey Release: Is the Biggest Barrier to AI Adoption Really 'Lack of Talent'? Flywheel Publishes '2026 Corporate AI Adoption Reality Survey' with 426 Large Enterprise Respondents. The True Obstacle to Sustainability is a 'Data Infrastructure That Cannot Measure Results'

Flywheel Inc. (Headquarters: Chiyoda-ku, Tokyo; CEO: Naoto Yokoyama; hereinafter 'Flywheel') has conducted and announced the results of the '2026 Corporate AI Adoption Reality Survey,' targeting 426 employees from large enterprises with 1,000 or more employees.

To date, Flywheel has conducted 'AI-Ready Assessments (maturity evaluations)' for over 400 companies. Building on this expertise, the company launched this survey to objectively capture the current state of AI adoption in enterprise organizations. The results show that while many companies identify 'lack of AI talent' as the biggest barrier to AI adoption, a detailed data analysis reveals that the true reason AI fails to take root is a structural issue: the inability to measure results and sustain improvement cycles. The 'barrier' companies face in AI adoption has already shifted from the 'introduction' phase to the 'operation and stabilization' phase.

Survey Results Slide (Partial)

Key Findings

The Gap Between Perceived Challenges and the Real Bottleneck

Although many companies cite 'lack of AI talent' as the primary obstacle, when AI adoption maturity was evaluated across six axes, the 'stabilization and improvement' phase—where evaluation metrics and improvement cycles are not established—recorded the lowest and most critical scores.

The Barrier of Data Quality and Return on Investment (ROI)

Many companies report difficulty in clearly seeing the 'results or ROI' from their AI initiatives. Underlying this is a data infrastructure challenge: the quality and reliability of data needed to support enterprise-wide AI use (e.g., RAG), or the preparation of unstructured data such as documents and logs, has not kept pace.

Perception Gaps by Role and Industry (Industry-Specific Barriers in Manufacturing)

Perceptions of AI adoption bottlenecks differ significantly between executives/decision-makers and frontline operators, as well as across industries such as manufacturing versus IT/software. In particular, manufacturing companies strongly perceive 'psychological resistance from the frontline' and 'security concerns' as bottlenecks—more so than IT and software companies—revealing unique challenges in operational integration.

Concerns About Future Gaps Driven by AI Adoption Levels

A majority of companies responded that AI adoption levels will create 'decisive or significant differences' in competitiveness within their industries over the next 3–5 years. This gap is expected to manifest not only in 'operational efficiency' but also as a disparity in companies' overall 'data utilization capabilities'.

Insights from the Survey: What Is an 'AI-Ready Data Infrastructure' to Overcome Post-Implementation Barriers?

The survey results suggest that the true reason for the lack of progress in 'stabilization and improvement' is not a lack of effort from frontline teams, but rather the absence of a data foundation ('infrastructure') necessary to objectively measure AI outcomes. To properly evaluate and improve AI output quality and ROI, reliable underlying data is essential. The key to overcoming the barriers of the 'operation phase' after implementation lies in building an 'AI-Ready data infrastructure'—one that supports three critical design principles: 'measuring' AI results, 'closing' improvement loops, and 'storing' insights in a reusable manner.

We will host an online webinar to explain deeper analysis results, including the 'differences in bottlenecks by role and industry' and 'specific numerical data from maturity assessments' mentioned in this release. Participating companies will also learn approaches to assess their own organization's current status.

Title: Uncovering Your Organization's Current State from 400 AI-Ready Assessments and the AI Implementation Roadmap for Maximizing Success

Date and Time:

【1】Live Streaming (Real-time)

June 19, 2026 (Fri) 13:00–14:00

【2】Archive Streaming (Recorded)

June 24, 2026 (Wed) 12:00–14:00

June 26, 2026 (Fri) 12:00–14:00

Location: Online (Zoom) / Free of charge

Content to be presented and explained in this webinar:

Detailed scores on 'Strategy & Policy,' 'Infrastructure,' 'Governance,' etc., derived from 426 respondents' data

Key differences in AI adoption between manufacturing and IT/software industries

The true nature of the perception gap between executives (decision-makers) and frontline staff

The '4 Essential Requirements' for linking AI to business outcomes and the steps to achieve them

Details and Registration: https://conata.flywheel.jp/20260619webinar

※Information on how to access the full survey report will be provided during the webinar and on our official website.

CEO Statement

Naoto Yokoyama, Representative Director and CEO

'This survey reveals the structural reality behind AI adoption challenges—issues that cannot be fully explained by the term 'talent shortage.' Many companies are stalling at the operation and stabilization stage because the foundational data required to measure AI outcomes, drive improvements, and accumulate knowledge is not in place. Under our mission of 'turning data into human energy,' we will strongly support enterprises from the data infrastructure side to help them overcome this 'post-implementation barrier' and connect AI to tangible business benefits.'

Survey Overview

Survey Name: 2026 AI Adoption Reality Survey Interpreted Through Six Maturity Axes

Target: Employees working at companies with 1,000 or more employees

Sample Size: 426 respondents (n=331 for some questions)

Method: Online survey

Survey Period: March 2026

Conducted by: Flywheel Inc.

【 About Flywheel Inc. 】

Flywheel is a professional services firm that drives business problem-solving and revenue growth through data utilization. By providing tailored consulting (Professional Service) and system implementation, we deliver visible results in a short timeframe. Through initiatives that smoothly and rapidly close the PDCA cycle of data utilization, and through our data utilization platform Conata® and consulting services, we promote secure and safe data utilization that maximizes protection of personal information and privacy. We contribute to building a society where businesses can effectively leverage data and solve real-world challenges.