This is article six in a six-part series exploring new research findings from HCLTech and Raconteur. You can read the rest of the article series in ‘The Blueprint for AI Leadership’ here or download the full report here.
AI has moved beyond experimentation. Across industries, organisations are moving from pilots and proofs of concept to confronting a more fundamental challenge: scaling AI into a core business capability.
At this stage, the dividing line is no longer adoption – it’s execution. Most organisations have proven that AI can deliver value in isolated use cases. Far fewer have translated that potential into sustained, enterprise-wide impact.
This gap is not explained by investment or access to technology alone. It reflects how organisations are building: the extent to which leadership, workforce capability, data and architecture are aligned around a common objective. As Pawan Vadapalli, CVP and global head, digital business services at HCLTech, puts it: “The difference between organisations that scale AI and those trapped in perpetual piloting comes down to a fundamental orientation. Leaders treat AI as a business transformation, while Followers treat it as a technology experiment.”
For leadership, the question is no longer whether to invest in AI, but how to make it deliver. And to that end, how an organisation answers four key questions defines the difference between experimentation and enterprise capability.
Who are AI Leaders & Followers?
AI Leaders, who represent 18% of the sample, are maximising ROI not simply by using AI to reduce costs, but by applying it to drive growth, innovation and improved customer experience.
AI Followers, by contrast, represent 60% of the sample. While they are realising returns from AI investments across business functions, they significantly trail Leaders in the higher-value use cases that are reshaping operating models and competitive advantage.
1. Are you connecting AI to measurable business outcomes?
The organisations that scale AI most effectively do not treat it as a standalone innovation agenda. They anchor it in clearly defined business outcomes, aligning use cases to the metrics the business already prioritises, such as revenue growth, cost efficiency, customer retention and operational performance.
Vadapalli says that AI initiatives must be tied directly to metrics the organisation already cares about: “When AI is framed as a business initiative, as a driver of revenue growth, margin improvement, customer satisfaction or competitive differentiation, it competes in a different arena and commands more attention and resources.”
This is often the point at which initiatives begin to stall. Where value is loosely defined or disconnected from core business priorities, AI remains confined to experimentation. Activity increases, but impact remains limited.
“What differentiates Leaders is not just their ability to deploy AI, but their ability to tie it directly to business outcomes,” says Vadapalli. “Followers, on the other hand, too often scope initiatives narrowly and execute them in relative isolation. There’s no connective tissue linking experiments to a coherent capability or outcome.”
This discipline changes how AI is deployed. Instead of being layered onto the business, it becomes embedded within the enterprise applications and workflows where value is created. The focus shifts from what AI can do to what it should deliver.
The conclusion is clear. Organisations should narrow their focus to a small number of high-impact use cases and define success in terms that the business already measures. AI initiatives that are not tied to clear financial or operational outcomes should be deprioritised in favor of those that can demonstrate tangible value.
2. Is cultural debt preventing your workforce from scaling AI?
Clarity determines where AI should be applied. Workforce capability determines whether it can be sustained.
Many organisations are confident that they can prepare their people for AI-driven change. Far fewer have translated that confidence into comprehensive, organisation-wide retraining strategies, redesigned roles or the operating structures needed to support adoption at scale. The result is a persistent gap between belief and capability.
This gap is not simply a skills issue. It also reflects a form of cultural debt: deeply embedded ways of working built for stability, control and predictable execution, but poorly suited to the experimentation, adaptation and human-machine collaboration that AI requires. In many organisations, roles remain too rigid, learning remains too episodic and performance systems continue to reward existing ways of working even as leaders ask employees to adopt new ones.
For Leaders, workforce enablement is not treated as a periodic HR initiative. It is embedded into how work is organised, how performance is evaluated and how employees are encouraged to experiment with AI in real workflows. They recognise that AI literacy is not simply a technical training challenge, but a work design challenge. People need to understand not only how to use AI tools, but how their roles, decisions and responsibilities change when those tools become part of day-to-day work.
Followers, by contrast, often mistake confidence for readiness. They may recognise the importance of upskilling, but their efforts remain limited in scope, disconnected from incentives or detached from how work actually gets done. In these environments, AI adoption is too easily framed as an employee resistance problem, when the deeper constraints are often the absence of development infrastructure, leadership alignment and role clarity.
“Confidence in a rapidly evolving domain like AI does not come from isolated pilots or one-off training,” says Vadapalli. “It comes from embedding continuous learning into how the organisation operates and from building systems people understand well enough to trust. You cannot train people once and consider them prepared.”
This distinction will become more important as autonomous AI expands. Employees will need to do more than operate systems. They will need to interrogate outputs, challenge decisions, identify anomalies and exercise judgment where human context still matters. These capabilities will not emerge organically. They must be deliberately built into roles, workflows and learning systems.
This is where workforce strategy becomes inseparable from AI strategy. Organisations should define what each role looks like when augmented by AI, align upskilling with performance and work design and confront the cultural debt that prevents experimentation, learning and adaptation from becoming part of everyday work. Without this, AI may be implemented, but it will not become a sustained enterprise capability.
3. Is your data foundation enabling AI or exposing its limits?
Data remains the most widely recognised dependency in AI, but also one of the most persistently underestimated.
Many organisations enter AI initiatives assuming their data environments are sufficient. It is only when they attempt to scale that gaps in quality, accessibility and governance become visible. In this sense, AI does not solve data problems – it exposes them.
“There’s often a significant gap between what leaders believe about their data estate and what AI teams discover when they try to use it,” says Vadapalli. “That gap becomes a constraint very quickly.”
Those successfully scaling AI take a different approach. They treat data as an enterprise asset rather than a byproduct of systems and processes. They invest in governance, standardisation and accessibility over time, creating environments where data can be reliably reused across functions.
The distinction is not simply technical–it’s personal. In environments where data is fragmented or closely held within functions, AI remains localised. Where it is shared, trusted and actively managed, AI is far more likely to scale. The priority for organisations should not be pursuing more data but making existing data more usable. They should focus on strengthening governance, standardisation and accessibility, treating data as a shared enterprise asset rather than a collection of siloed resources.
4. Is your architecture empowering AI-driven transformation or holding it back?
Even as data improves, architecture often lies in wait as the next constraint. AI places fundamentally different demands on enterprise systems. It requires interoperability across applications, fluid data movement across domains and the ability to embed intelligence into workflows that were never designed to support it. Legacy environments, often optimised for stability and efficiency, struggle to meet these requirements.
In many cases, these constraints reflect accumulated technical debt: legacy systems, fragmented architectures and integration patterns that were never designed for AI at scale. The result is a growing disconnect between ambition and execution. Organisations can develop AI capabilities at the surface, but struggle to embed them into core operations.
“Most enterprises can buy tools and hire talent but scaling AI is not just about adding new capabilities,” says Vadapalli. “It’s about removing the constraints that prevent those capabilities from working at scale. Leaders must make hard trade-offs, fund the foundations before the big wins are visible and standardise patterns even when teams prefer autonomy and enforce governance without slowing innovation to a crawl.”
Simply put, modernisation is unavoidable, but also inherently difficult and costly. It requires new investment and new technologies, of course, but also coordinated changes to existing systems, processes and ways of working that have evolved over time.
“Discipline is what turns scattered AI activity into something cumulative,” adds Vadapalli. “Without it, organisations end up with pockets of AI excellence but an enterprise that still can’t scale.”
Organisations that are making progress take a more deliberate approach. They prioritise the areas where architectural change will unlock the greatest value, adopt more composable models and align modernisation efforts with business priorities. Without this, AI’s potential remains peripheral, capable of generating insight but unable to drive meaningful change.
Leaders should resist the temptation to layer AI onto existing systems without addressing underlying constraints. Instead, they should prioritise targeted modernisation and focus investment on the parts of the architecture that will unlock the greatest business value. Doing so sequences change in a way that supports both innovation and long-term scalability.
Turning ambition into advantage
Together, these questions point to a broader shift. The primary challenge in scaling AI is not related to your technology but to your operating model.
Though in reality, the barriers to scaling AI are rarely isolated. They reflect a combination of technical debt and cultural debt–legacy systems on one hand and legacy ways of working on the other–alongside fragmented data environments and misaligned operating models. Addressing one without the others rarely delivers sustained progress.
The organisations pulling ahead are not those with the most advanced models, but those that have aligned the conditions required to make those models effective. They connect AI to measurable outcomes. They build trust across the workforce. They strengthen data foundations rather than working around existing limitations. And they modernise architecture in ways that enable, rather than constrain, progress.
For Vadapalli, two priorities stand out. First, confront the cultural debt that limits AI adoption by redesigning roles, incentives and learning systems around AI-augmented work. Second, ensure leadership commitment is visible and sustained, not just stated, while putting humans at the centre of this AI-driven transformation.
These actions do not guarantee success. But they create the credibility and momentum on which broader transformation depends.
Ultimately, the question is no longer whether AI will reshape the enterprise. It is whether business leaders are prepared to reshape their organisations around it.
This is article six in a six-part series exploring new research findings from HCLTech and Raconteur. You can read the rest of the article series in ‘The Blueprint for AI Leadership’ here or download the full report here.
This is article six in a six-part series exploring new research findings from HCLTech and Raconteur. You can read the rest of the article series in 'The Blueprint for AI Leadership' here or download the full report here.
AI has moved beyond experimentation. Across industries, organisations are moving from pilots and proofs of concept to confronting a more fundamental challenge: scaling AI into a core business capability.
At this stage, the dividing line is no longer adoption – it’s execution. Most organisations have proven that AI can deliver value in isolated use cases. Far fewer have translated that potential into sustained, enterprise-wide impact.
This gap is not explained by investment or access to technology alone. It reflects how organisations are building: the extent to which leadership, workforce capability, data and architecture are aligned around a common objective. As Pawan Vadapalli, CVP and global head, digital business services at HCLTech, puts it: “The difference between organisations that scale AI and those trapped in perpetual piloting comes down to a fundamental orientation. Leaders treat AI as a business transformation, while Followers treat it as a technology experiment.”
