I dislike the term AI strategy. AI can be a powerful tool within an IT or business transformation strategy, but when it becomes the strategy itself, organizations risk losing sight of the outcomes they are trying to achieve.
Successful transformation depends on planning, meaningful measures of performance, and people who are willing to accept the change.
That distinction is becoming more important as AI capabilities advance.
Senior Legal Counsel, AI, InterSystems.
Organizations are understandably eager to adopt the technology and gain a competitive advantage.
The race is on, but lasting value will depend on asking a more disciplined question: where can AI create meaningful business impact, how should it be governed, and can it scale beyond initial experimentation?
For many organizations, that leap from experimentation to scale remains unresolved.
Recent McKinsey research found that most organizations remain in experimentation or pilot modes. As regulation adds ambiguity for those rushing to deploy AI solutions, strategic planning becomes critical.
AI needs business ownership
For any business transformation, stakeholder engagement is the first hurdle. The challenge is turning early interest and isolated pilots into initiatives that are scalable, well-governed, and tied to long-term enterprise outcomes.
For scalable AI enablement, a macro view is required. A lack of synergy between invested stakeholders is frequently why AI stalls at pilot phase. The strategy cannot be contained just within the IT department. Instead, collaboration across legal, compliance, technology, operations, and commercial teams is crucial, especially as AI risk becomes harder to separate from business risk.
With the ambiguity of AI regulation and geopolitical agendas, strategic partnership with legal teams is one aspect of this stakeholder engagement that is increasingly in demand, helping organizations respond to uncertainty with more agility built in from the start.
Once the right people are involved, the next step is to understand the business problem. During initial planning, business processes should be coherently mapped. Once the problems are identified, solutions can be found.
Those solutions may not involve AI at all, of course. It should be considered whether standard IT technology or another process would fix the problem first.
If AI is the right tool, it still needs to be introduced with a clear understanding of workflow processes. Otherwise, organizations risk throwing AI at problems in a rushed, bolt-on style.
Multiple AI use cases and vendors can exacerbate the inefficiencies that AI was supposed to fix and create more regulatory burden. Long-term value will come from identifying where AI can improve specific processes, with simplification not duplication, before deciding which opportunities are ready to scale.
This is where speed needs to be balanced with adaptability. Responsible AI and governance are building blocks for regulatory readiness and compliance, but they are also levers that allow organizations to shift gears when needed.
Measuring value beyond productivity
Once AI has been framed as part of a wider transformation strategy, organizations need to think carefully about how success will be measured. Too often, AI remains a technology initiative in a silo, assessed mainly through productivity gains or short-term cost savings. These metrics are useful but are not the full picture.
The real test is whether AI improves the quality and efficiency of work, reduces risk, strengthens business outputs, and promotes employee wellbeing and job satisfaction.
Many organizations are also grappling with clear KPIs, and tangible ROI. If internal AI adoption is inconsistent, and if organizations struggle to identify use cases that can scale rather than remain isolated pilots, the return will never be realized. People need to use AI tools for the benefits to be proven.
Even when adoption improves, early KPIs may not be immediately tangible. They may not appear quickly in financial reporting or obvious performance metrics.
Yet AI’s value can still be felt in the quality of work itself, reducing repetitive tasks and giving employees more capacity to focus on higher-value activity. That is why organizations need to look beyond the most obvious measures of value.
Over the longer term, AI deployment will facilitate a shift away from administrative or duplicated tasks, creating more time for strategic initiatives. Beyond the human element, though, organizations should measure AI against broader business outcomes such as decision quality, risk reduction, resilience, and customer impact.
The same broader view of value should shape how pilots are designed. Pilots are useful for testing proof of concept, refining guardrails, and encouraging cultural adoption. Yet they are often conducted in a closed setting, on synthetic data, with limited users and under supervision from the technology supplier.
As AI moves closer to scale, organizations need to revisit those early assumptions and assess whether the tool can still deliver value in real-world conditions. In short, lessons learned from pilots, including how costs and user behavior may shift at scale, need to be applied across wider adoption.
Trust will determine scale
As organizations move from experimentation to broader deployment, supplier strategy becomes critical. Even well-defined AI use cases can struggle to scale if the technology cannot be embedded into core systems and workflows. Once processes and problems are mapped, selecting the right supplier is where strategy starts to play out in practice.
Businesses that overlook suppliers that promote responsible AI and compliance will be at a competitive disadvantage here. They may waste money on solutions that will not work, will not scale, and may need to be reprocured later. They may also fail to meet regulatory obligations if they cannot source or record information such as data lineage or transparency around when users interact with AI tools.
Gartner has argued that AI value depends on business-aligned pilots, IT infrastructure readiness, and coordination between AI and business teams. Trust and transparency are therefore becoming the new currency, especially as the Thomson Reuters Foundation and UNESCO have warned of a widening transparency gap in corporate AI adoption.
For AI to scale, governance cannot be treated as an administrative burden. Effective AI governance covers use case risk classification, tool selection, accountability, security reviews, supplier due diligence, and ongoing monitoring. It provides the playbook for future expansion and helps organisation explain their systems externally.
That external trust needs to be matched internally. Scalable AI adoption will hinge on whether employees trust the system, understand how tools map against process reform, and feel that their own judgement still matters. If they do, they will use the tools, actualize the KPIs, and help refine deployment.
Long-term value and scalability from AI is less about speed, automation, and cost-saving. It comes from using the technology strategically, with governance and trust built in from the start. For organizations that recognize those nuances, AI stops being treated as a singular strategy and becomes part of business strategy.
This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.
The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit