Closing the AI trust gap: What leaders are seeing about trust in AI
Fujitsu Trust in AI Report
Article | 2026-6-25
10 minute read
Business leaders are excited by AI, but their governance is dragging them down. Trust can supercharge their operations. But how?
Trust in AI is emerging as a clear marker of good business performance, according to new research from Uvance Wayfinders, Fujitsu’s consulting business.
The study of business leaders highlights a link between those that have full confidence in their AI applications and better business performance on key metrics such as productivity, product development and commercial success.
Those with the fullest confidence in their AI solutions are 74% more likely than the rest to be meeting their financial goals, and 60% more likely to be improving productivity and internal efficiencies.
Unfortunately, most companies can’t trust their AI: nearly four in ten business leaders in the research (38%) don’t have faith in their AI input data, saying it’s fragmented, unstructured and has quality issues. And 39% say their personal reputation and role will be at stake if an AI project goes wrong because of insecure, unethical or untrustworthy input data.
In this report, we examine how good data hygiene, investing in skills and prioritizing governance can help businesses close the AI trust gap and begin to reap the benefits of trusted AI.
Business leaders see huge potential in AI
It’s difficult to overestimate AI’s impact on businesses. As exciting pilots turn into scale-up opportunities, organizations are making major investments in the technology. According to Gartner, worldwide spending on AI will reach $2.52 trillion in 2026 – up 44% annually – and next year it will rise by another third to $3.34 trillion (*1).
Business leaders in the Uvance Wayfinders research agree about AI’s potential. Seven in ten (69%) say that AI will be the biggest game-changer for businesses since the rise of the internet.
And it’s not just hype. They believe that AI will have a genuine impact on their business: 60% expect their use of AI to lead to significant cost savings in the next two years, and 50% expect it to underpin a major new product line or refresh over the same period. Scaled across the industry, these results would amount to more than incremental change – they would signal a major transformation.
AI innovation is speeding ahead
Given the hype, it’s no surprise that businesses are pressing ahead with innovation. Almost every area of their businesses is becoming AI-enabled.
Our data shows that AI is becoming pervasive, and that only a small minority of business leaders have no AI projects planned across the use cases we asked about. Over time, AI is going to touch every part of the business.
Organizations are now using AI extensively
Organizations have already implemented the more accessible use cases, and are starting to explore more technical uses for AI. For instance, 60% of business leaders say that projects are currently live in document processing and information-extraction tasks, while 58% use AI for predictive analytics and forecasting. And projects are live in even more complex deployments, such as AI-assisted software development and model-monitoring and governance tools.
The AI trust gap is limiting potential
As AI moves into customer-facing, revenue-critical and compliance-sensitive areas, trust becomes a prerequisite – not an optional enhancement. It’s not just about whether AI systems function correctly.
It’s about decision assurance and operational resilience: organizations must be able to explain how decisions are made, demonstrate control over data across jurisdictions, secure systems that are increasingly interconnected, and sustain performance as deployments scale.
But progress is constrained by a major challenge: the AI trust gap.
Business leaders have doubts about the accuracy and reliability of AI tools. And because of the size of investments and the sensitivity of the reams of data AI applications deal with, they can’t ignore those doubts.
What’s causing this distrust? Nearly four in ten business leaders (38%) don’t have faith in their input data. They believe it’s fragmented and unstructured – and is therefore not fit to support AI. This is a serious concern, because high-quality data is at the center of ethical and compliant AI use cases.
This is a foundational issue that’s causing serious concern among business leaders: 77% say they would never launch an AI application without making sure it’s completely trustworthy. But how long would it take for business leaders to build enough trust in their systems and data?
Six in ten, meanwhile, have concerns about AI agents and believe them to be less trustworthy than non-autonomous applications. They’re right to be cautious about this: 39% of business leaders say their personal reputation and role are at stake if an AI project goes wrong because of insecure, unethical or untrustworthy input data.
Poor governance is widening the AI trust gap
It’s no surprise that business leaders are failing to build trust: our data shows that organizations are nowhere near prepared enough when it comes to the governance that underpins AI applications. Their decision assurance, operational resilience and risk posture are under threat.
In crucial areas, organizations’ AI governance strategies aren’t advanced enough
Only 13% of business leaders say their security and sovereignty controls are “best-in-class,” and 12% say the same about their data readiness. This is a big problem, because these are fundamental areas where only the best will do.
Basic, high-level or draft policies, which appear to make up most organizations’ guidelines, won’t create trusted AI outputs. This can affect how customers see the organization, and damage their trust.
Taken together, the AI adoption data and governance data create a problematic picture that exemplifies the AI trust gap:
-
51% of business leaders have live projects for internal content generation, but 47% have only basic explainability practices in place. This suggests that the source of the information that powers the content, and how it has been applied, is a mystery to many organizations.
Considering how sensitive many internal materials are, this is a serious shortcoming. -
36% of business leaders say their organizations have live projects for AI
assisted software development. Separately, 53% of business leaders report having only basic data readiness practices in place. This suggests that the data underpinning software development might not be a strong enough foundation, and could cause problems as the organization embeds the software into its operations. - 54% of business leaders say they have live projects for customer service implementation, but 30% have only basic security and sovereignty practices in place. This suggests that customer data might not be handled appropriately, which introduces significant reputational and regulatory risks and can undermine operational resilience.
So there’s a wide gap between business leaders’ expectations of AI and their ability to bring these to life with solid, trustworthy governance practices. What’s stopping them from improving those practices?
The good news is that dealing with these barriers is a top priority for business leaders in the coming 12 months.
The top barriers preventing business leaders from building trusted AI
Business leaders are prioritizing data quality in the year ahead
High-trust organizations outperform
Organizations are right to make overcoming governance barriers a priority. Our research shows that companies that have the most trust in their AI systems perform better than the ones with less trust.
This suggests that building trusted AI systems could be good business. While we can’t claim direct causation between the two, the link between trusted AI and better business outcomes is notable.
The data shows that organizations with full confidence in their AI applications outperform the ones that have less confidence across important performance metrics: productivity, developing new products, generating ideas and meeting financial goals.
For companies that are finding it hard to close the AI trust gap, this should be an incentive to keep going.
Companies that are performing well on the business outcomes
How to close the AI trust gap
The question isn’t whether organizations will deploy AI – that’s now settled. The question is whether they can deploy it on a strong foundation of trust and accountability.
Here’s how to get there:
1. Fix the foundations
Trust doesn’t begin in the AI application; it starts in the underlying data and how it gets managed. Fragmented, unstructured data weakens explainability, limits scalability, fractures operational resilience and ultimately creates an AI trust gap. Overcome this by building in control and management from the bottom up. Consolidating data environments, establishing clear ownership and embedding continuous data quality monitoring help companies to do this.
2. Turn governance principles into operational reality
Governance must move beyond rough guidelines and policy documents and operate at the same speed as AI innovation. As the technology’s autonomy increases, oversight protects both performance and reputation. Embed explainability and traceability mechanisms into workflows to strengthen decision assurance, define risk thresholds and escalation pathways, and deploy model-monitoring tools to manage bias, security and compliance risks. Evolving governance on an ongoing basis is essential to long-term success.
3. Upskill employees to demystify AI
When employees understand AI, they’re more likely to trust it and use it responsibly.
Organizations that prioritize AI upskilling will find that trust becomes embedded by design, rather than imposed by policy. Focus on role-specific AI training, provide insight on how models work and empower employees to challenge outputs. This will boost their confidence and make them more responsible users.
Speed matters when it comes to AI adoption, but not at the expense of trust. Companies that close the AI trust gap will be the ones that turn potential into real business outcomes.
About the research
In February 2026, a Uvance Wayfinders commissioned survey conducted by FT Longitude surveyed 400 senior business leaders based in Australia, Japan, the UK and the US. Leaders represented technology and IT, finance, strategy and operations equally, and were from companies across the following sectors: financial services; manufacturing; energy, resources and utilities; logistics and supply chain; healthcare and life sciences; retail and consumer goods; the public sector, government and defense; technology and telecommunications; and professional services. Half (53%) of companies had between 1,000 and 4,999 employees, 38% had between 5,000 and 49,999 employees, and 10% had more than 50,000 employees. Percentages throughout this report may not sum precisely due to rounding.
Related Information
Data sovereignty in the AI era: from risk management to strategic advantage
Find Uvance Wayfinders in your region
Contact us