As Chief Care Officer at CareBrain, I meet providers who feel they must pick the “one AI system” that will transform everything. In practice, AI success has far less to do with choosing a platform and far more to do with how people adopt it – what problems it genuinely solves this month, how workflows shift around it, and whether staff feel confident using it in real care settings.

Across sectors, only one in four AI initiatives has delivered the returns organisations expected over the past three years. The emerging evidence suggests the barrier isn’t the technology itself. It’s the environment it’s being introduced into. Many organisations are still operating within 20th-century structures built for control and uniformity, then expecting modern AI to thrive inside them.

Commentary in Fast Company and research referenced by Outpoll describe a different approach: redesigning around people first. Some call this building a “Human OS” – a human-centred operating model that supports iteration, feedback and adaptation. Where that redesign happens, AI is far more likely to generate measurable and sustainable return on investment.

Leadership plays a central role in this shift. Reporting from ZDNET highlights how some organisations are moving beyond the generic “Chief AI Officer” title and instead appointing a Director of AI Productivity. The distinction is practical. This role focuses on bridging data, IT and frontline operations, and on driving real adoption of the enterprise-grade AI tools organisations are already funding. The emphasis is not on experimentation alone, but on consistent usage and measurable productivity gains. Without adoption, return on investment remains theoretical.

In health and social care, this matters even more. When leaders do not provide clear direction, approved platforms and visible early wins, staff will understandably look for ways to reduce administrative pressure. That can mean turning to free, unsanctioned AI tools to draft reports or summarise meetings. While the intention is efficiency, the risks around data protection, governance and quality are significant. A clearly identified adoption lead – whether titled Director of AI Productivity or something similar – creates guardrails, enables secure tools and keeps attention focused on care outcomes rather than novelty.

For business leaders asking how to ensure AI investment succeeds, the starting point is rarely a large-scale transformation programme. It is far smaller and more practical. Pilot AI in areas where the benefit is obvious and measurable: compliance auditing, meeting summaries, staff supervision documentation or point-of-care coaching. Early wins create confidence and provide internal evidence about what should scale next. This is how an AI strategy becomes grounded in operational reality rather than ambition alone, a point reinforced in discussions covered by Fast Company and Outpoll.

Alongside this, workflows and training need to evolve. AI should reduce cognitive load and administrative burden, not add another layer of complexity. When staff experience tangible time savings and clearer documentation, adoption grows naturally. When the opposite happens, resistance is predictable.

Finally, ownership must be explicit. Whether the title is CAIO or Director of AI Productivity, someone needs the authority to connect IT, data and frontline teams, monitor uptake, review impact and adjust implementation regularly. Adoption improves when responsibility is visible and continuous.

Across all of this, the pattern is consistent. Technology alone does not create value. Value emerges when people use it confidently, safely and consistently within modernised ways of working. Organisations that empower teams, update workflows and scale from small, well-managed pilots tend to see AI deliver measurable and sustainable impact – for staff and for the people they support.

If AI investment is on your agenda, measure adoption as carefully as you measure spend. In care especially, adoption is what determines whether AI strengthens practice or simply adds another unused system.

For organisations exploring practical, governed AI adoption in health and social care, more insights and implementation thinking can be found at carebrain.ai, grounded in real-world care delivery rather than theory.