Why Most Businesses Are Still Struggling to Win with AI
While 86% of global executives call AI core to their strategy, new reports from Kyndryl, Qlik and S&P Global reveal that complexity, cultural inertia, and capability gaps are keeping organisations from unlocking its true potential.
AI is everywhere. It’s the cornerstone of transformation roadmaps, the centrepiece of boardroom conversations, and increasingly, the north star of enterprise innovation.
A recent Qlik study revealed that 86% of senior executives say AI is now central to their organisation’s business strategy. Yet, only a small fraction are seeing the meaningful business outcomes they hoped for.
A parallel report by Kyndryl paints a similar picture. Despite the enthusiasm, most businesses are still stuck in the early phases of AI maturity. Only 5% of organisations are considered ‘AI Pacesetters’— those that successfully use AI at scale and see significant returns.
So what’s going wrong?
The AI Execution Gap Is Real
Both reports point to a sobering truth: strategy alone isn’t enough. There’s a wide and growing gap between AI ambition and AI execution, and it’s costing companies time, money, and competitive advantage.
“Organisations clearly recognise that merely investing in AI is insufficient; what matters now is delivering tangible outcomes. Yet, as our research underscores, the road to production AI remains blocked by persistent hurdles—cost, complexity, and data fragmentation,” said Mike Capone, CEO of Qlik.
Closing the AI execution gap requires more than aspiration—it demands practical solutions that simplify data integration, ensure governance, and empower better decision-making.
This pressure is even more intense with generative AI dominating leadership agendas. The pace of genAI evolution amplifies organisational anxiety and widens the gap between intent and capability.
Eric Hanselman, Chief Analyst at S&P Global Market Intelligence, said, “The fast-evolving GenAI landscape pressures enterprises to move swiftly, sometimes sacrificing caution as they strive to stay competitive. Many are deploying GenAI tools before fully understanding their implications, especially with the surge of SaaS platforms embedding genAI capabilities.”
Recent research from S&P, ‘The 2025 Thales Data Threat Report’ revealed that nearly 70% of organisations consider the fast pace of generative AI development the leading challenge tied to AI adoption, followed by concerns over integrity (64%) and trustworthiness (57%).
Enterprises are leveraging AI to accelerate product development, enhance CXs, improve training, speed drug discovery, and optimise operations. However, the rapid adoption of genAI introduces complex challenges that organisations must navigate carefully.
The Five Core Blockers Include:
1. Workforce Inertia and Fear
- Kyndryl found that 71% of leaders believe their workforce isn’t ready to adopt AI.
- 45% say there’s active resistance or even fear of job displacement due to AI.
2. Talent and Skills Shortages
- Over half of the organisations surveyed (51%) admit they lack the necessary AI-skilled talent to scale effectively.
- Many are not investing fast enough in reskilling or change management.
3. Data Complexity
- Qlik’s report shows most organisations are bogged down by fragmented data systems, legacy infrastructure, and inconsistent governance models.
- Nearly 80% of respondents say these issues are their biggest barriers to realising AI’s full potential.
4. Leadership Disconnects
- Decision-makers often approve AI investments without aligning execution with on-ground operational realities.
- Only a small portion of leaders feel confident that they have visibility into how AI projects are progressing.
“A profound but succinct insight I heard is that ‘The hard stuff is the soft stuff.’” Successful transformation is not about upgrading technology; it is about managing people’s reaction
to technology and change. This is even more challenging now, with AI evolving at an unprecedented pace. Many fear obsolescence, irrelevance, or the loss of their livelihoods,” says Patricia Mulles, Director and Global Head of Partnerships at She Loves Data and Founder of M & O Infinite Library.
5. Isolated Use Cases
- Many companies treat AI like a set of standalone pilot projects instead of weaving it into the broader operating model.
- This results in fragmented impact and little measurable ROI.
Data Fluency: The Missing Layer
Qlik’s report hammers home one overlooked truth: without a data-literate workforce, AI investments are doomed to underperform. Companies need to think beyond models and infrastructure. It’s about ensuring teams at every level can ask the right questions, interpret results, and make decisions powered by insights.
One of the right questions to ask is this: Does my customer trust me with AI?
According to Kate Parker, President at Transcend, user consent may not have the same allure as flashy AI tech, but it is the linchpin of the digital economy’s next phase. Trust isn't just an internal
issue. It also shapes how customers perceive and engage with AI-driven experiences.
“AI initiatives will falter without the right infrastructure to honour user choice and govern data. Companies that don’t prioritise consent will alienate customers and rack up fines, undermining their AI ambitions and their bottom lines,” Kate adds.
AI Pacesetters: What the Leaders Are Doing Differently
Despite these obstacles, there are organisations that are getting it right. Kyndryl categorises them as AI Pacesetters and their success patterns are revealing:
- They operationalise change management.
Pacesetters are 3x more likely to have a structured plan for managing organisational change tied to AI initiatives. - They build employee trust.
They report 29% fewer employee concerns about AI replacing jobs, focusing instead on how it augments roles. - They prioritise skilling.
These organisations are 67% more likely to accurately assess and build the skills their employees need to work with AI. - They use AI for more than automation.
While lagging companies still associate AI mostly with efficiency and cost-cutting, Pacesetters leverage it to enhance customer experience, enable personalisation, and support strategic decision-making.
Here’s a real case study. IBM collaborated with Scuderia Ferrari to increase fan engagement through a reimagined Scuderia Ferrari app powered by IBM’s watsonx AI platform. The app processed over a million data points per second during races, transforming this complexity into immersive experiences for Ferrari’s 400 million fans worldwide.
It delivered personalised content, insights, and interactions, while also improving internal workflows and scaling content creation. With this well-thought of partnership, the brand delivered scalable, high-value user experience while enhancing internal productivity.
To be a Pacesetter, here’s where to focus:
- Start with people, not platforms.
Educate and empower your teams to understand, use, and trust AI. - Fix your data.
AI can’t thrive in silos or chaos. Clean, structured, accessible data is a non-negotiable foundation. - Align intent with execution.
Strategic vision means nothing without frontline enablement. - Measure what matters.
Shift from vanity metrics to impact-based indicators—customer experience, operational agility, and workforce engagement.
“The encouraging news is that businesses that can get alignment at the top are not only marching in the same direction, but are seeing benefits,” said Kim Basile, Chief Information Officer at Kyndryl. “This work isn’t easy, but aligning technology strategies with broader business goals is the top action leaders need to take to fully benefit from AI.”
ALSO READ: 25 Martech Innovators Injecting Agentic AI into the Suite
