Businesses Need Better Support for AI: Survey Shows Gaps in Data, Talent, and Policy

The OECD, BCG, and INSEAD study reveals that while many firms actively adopt AI, they face challenges around data access, regulation, and talent. The report urges streamlined public support, clearer accountability, and stronger collaboration to accelerate responsible AI integration.

A study jointly produced by the OECD, Boston Consulting Group (BCG), and INSEAD Business School delivers a thorough and timely account of how businesses are integrating artificial intelligence across the G7 countries and Brazil. Drawing from a policy-oriented survey of 840 enterprises and in-depth interviews with both private and public sector stakeholders, the report examines how AI is reshaping decision-making, innovation, and productivity. What emerges is a complex picture of promise and challenge. Even firms with advanced AI systems continue to search for clearer regulatory pathways, more accessible public data, and practical guidance for implementation. These insights not only speak to the dynamism of AI technologies but also to the institutional lag that can slow their impact. The research underscores that successful AI integration requires not just cutting-edge tools but also well-functioning ecosystems, spanning data access, regulation, funding, and education.

Public Services Are Valued, But Gaps Persist

Public services designed to support AI adoption received mostly positive reviews from the surveyed firms. About 75% of enterprises in the manufacturing sector and 69% in the ICT sector said they relied on government-provided services such as economic data, market intelligence, regulatory updates, and compliance support. These tools were seen as valuable for strategic planning, business analytics, and market positioning. Yet, despite this appreciation, businesses expressed clear frustration with the fragmented nature of support infrastructure. Many respondents noted that identifying the right public agency or service often required unnecessary time and effort. There were repeated calls for the establishment of a centralized public platform or “AI hub” that could offer a one-stop interface for data access, vendor directories, funding opportunities, and regulatory resources. The lack of such integration was viewed as a serious barrier, especially for small and medium-sized enterprises (SMEs) with limited internal bandwidth to navigate complex government systems.

Data Access and International Rules Are a Bottleneck

One of the most critical constraints identified was the challenge of accessing clean, reliable, and up-to-date data. Public data repositories, while valuable in theory, were often outdated or poorly documented in practice. In many cases, enterprises had to spend considerable time validating the currency and integrity of datasets, which compromised the speed and accuracy of AI model development. Furthermore, the legal complexity around international data flows was flagged as a growing concern. The lack of alignment between privacy laws in different jurisdictions makes it hard for companies operating across borders to confidently share and process data. While frameworks like the EU’s General Data Protection Regulation (GDPR) are widely acknowledged as useful standards, firms emphasized the urgent need for more globally harmonized legal architectures. Interviewees recommended developing standardized protocols for data merging, sharing, and usage, particularly for SMEs working with international partners.

Collaboration and Talent: The Double Engine of AI Innovation

Collaboration between private firms and public research institutions emerged as another vital area for AI development. Most businesses acknowledged the value of such partnerships for talent access, technical validation, and innovation. Yet, the bureaucratic processes for setting up these collaborations, particularly those involving public funding, were often cited as overly complex. Enterprises advocated for simplified application procedures, clearer evaluation metrics, and better communication loops between funders and applicants. Additionally, they stressed the need for model contracts and standard agreements to speed up the legal and financial arrangements required for joint R&D. Talent development also came into focus. A staggering 86% of firms said public investment in retraining and lifelong learning for AI practitioners would be “very helpful.” Similarly, 82% endorsed more funding for university-level and vocational education in AI-related fields. Beyond technical upskilling, some firms even suggested increasing digital literacy among public officials to ensure better institutional support for AI deployment.

Regulation, Vendor Guidance, and the Path Forward

The regulatory environment for AI remains underdeveloped in many jurisdictions, and this uncertainty is affecting enterprise confidence. About 92% of respondents expressed strong support for regulations that clearly assign accountability in AI-driven systems, especially when such systems interact with customers or make autonomous decisions. Enterprises want regulators to go beyond general principles and offer sector-specific guidance. The idea of vendor certification also gained traction, particularly as companies, especially SMEs, struggle to differentiate between marketing hype and proven solutions in the crowded AI marketplace. A neutral, government-endorsed checklist or rating system was suggested as a way to bring transparency and assurance to the vendor selection process. Meanwhile, more traditional forms of support, such as tax incentives for R&D, investment in cloud infrastructure, and open administrative datasets, were also welcomed, reinforcing the idea that AI advancement depends as much on enabling infrastructure as on software algorithms.

A Framework for Future Policy

The landmark study highlights the multifaceted nature of AI adoption in firms: it is not merely a matter of purchasing software or hiring engineers, but of embedding intelligent systems into complex organisational, regulatory, and economic contexts. The findings suggest that while businesses are eager and increasingly capable of implementing AI, they need more coherent and accessible support systems. Policy makers are encouraged to act decisively, streamlining public services, simplifying funding mechanisms, clarifying regulatory expectations, and investing in both human and technological infrastructure. As AI becomes more embedded in economic activity, governments that provide thoughtful, user-centric support systems will empower their business sectors to innovate, compete, and grow in a rapidly changing world. The partnership of OECD, BCG, and INSEAD has not only diagnosed the challenges but has also laid out a pragmatic roadmap for the future.