Artificial intelligence is already reshaping how healthcare is planned, delivered and evaluated, from diagnostic support and treatment personalisation to workflow optimisation. A 2026 critical reflection published in npj Digital Medicine applies McCance and McCormack’s Person-Centred Practice Framework to the integration of AI in care. The framework links macro context, professional prerequisites, the practice environment and person-centred processes with the achievement of person-centred outcomes. Through this lens, AI carries a dual potential. It can support earlier detection, better use of resources, personalised planning and more responsive services. It can also narrow care around measurable outputs, weaken professional judgement, intensify surveillance and reduce individuals to data profiles. The central challenge is therefore not adoption alone, but implementation that protects dignity, relationship, agency and ethical accountability.
Policy and Governance Need Human Values
The macro context for AI in healthcare includes policy frameworks, strategic leadership, workforce development and wider system priorities. AI adoption is often driven by cost reduction, workforce optimisation, predictive analytics, interoperability and diagnostic precision. Policy examples include the UK plan, the European Commission approach, the European Union Artificial Intelligence Act, the United States AI risk management framework and China’s generative AI measures. These frameworks reflect growing strategic interest in AI but also reveal a strong emphasis on scalability, regulatory conformity and technical performance.
AI can support more equitable and preventive services when embedded in public health programmes, population monitoring and resource allocation workflows. Pattern detection may help identify underserved groups, emerging risks and opportunities for early intervention. However, measurable data can shape what becomes visible to the system. Experiences or populations outside dominant data structures may remain unrecognised, while predictive tools may reproduce historical underrepresentation or structural bias. Continuous risk profiling can also blur the boundary between support and surveillance.
Must Read: A Human-Centred Approach to Scaling AI
Person-centred AI governance requires more than cybersecurity and data protection. National strategies need explicit commitments to dignity, empathy, autonomy and participation. Co-creation with people receiving care, caregivers and health professionals can connect AI design with real care experiences. Oversight also needs algorithmic audits, interdisciplinary ethical review, attention to bias and feedback channels that allow concerns to inform technical refinement and organisational learning.
Professional Competence Must Evolve
The prerequisite domain of the Person-Centred Practice Framework focuses on the attributes healthcare practitioners need for person-centred care. These include professional competence, interpersonal skills, commitment to the job, clarity of beliefs and values and knowing self. AI can strengthen competence by providing access to clinical data, guidance, decision support and simulation-based education.
The same tools create new professional demands. Clinicians need the ability to interpret algorithmic outputs, judge reliability and combine machine-generated recommendations with clinical experience and contextual knowledge. Informatics competencies, digital ethics, critical appraisal and awareness of algorithmic limitations become part of professional practice. Without these capacities, practitioners may over-rely on automated outputs or dismiss useful insights because of mistrust.
AI-supported individual medication planning illustrates the tension. Systems can combine comorbidities, therapies, lifestyle factors and stated goals to suggest tailored regimens, flag trade-offs and identify risks. These outputs can structure discussion and make uncertainty more visible. They can also undermine person-centred care when they optimise mainly for adherence or cost, generate alert burden or encourage automation bias. The person’s preferences must remain determinative.
Interpersonal skill also needs protection. AI can intrude on consultation dynamics if screens, dashboards or recommendations reduce eye contact, listening and authentic engagement. Reflective practice helps practitioners examine how technology affects communication, empathy and ethical judgement. Knowing self becomes a safeguard against dehumanisation, requiring awareness of personal attitudes to technology, implicit biases and algorithmic authority.
Care Environments Shape AI Impact
The practice environment includes skill mix, staff relationships, organisational systems, power sharing, innovation, risk taking and physical space. AI can influence workload distribution through scheduling, resource allocation and task automation. These functions may reduce administrative burden and create more time for person-centred interactions. They may also reinforce inequities in staffing hierarchies, role expectations and workload if efficiency dominates design and governance.
Human oversight must involve real authority. In highly optimised systems, clinicians or managers may remain formally involved while lacking meaningful ability to pause, amend or reject algorithmic outputs. Blurred accountability between clinical leadership and information technology can make overrides risky. Several operational pressures can weaken human control. Person-centred practice requires interruptibility, contestability and institutional support for ethically motivated overrides.
AI adoption also depends on culture. Tools imposed without preparation, consultation or adaptation can disrupt workflows and team cohesion. Frontline workers therefore need involvement in selecting, designing and evaluating technologies. Supportive systems must provide infrastructure, training, equitable access to AI resources and safe forums for raising concerns.
Power should not concentrate in algorithms or technology vendors. Clinicians, service users and families need routes to co-govern tools that affect care. Ethics committees and advisory panels with diverse stakeholders can help maintain accountability. Physical environments also matter. Shared screen viewing can support transparency, while screen-centred layouts may fragment communication. Less intrusive alerts can help preserve relational presence.
AI can strengthen person-centred practice when it supports insight, access, shared decision-making and time for human connection. It can weaken that practice when technical performance, productivity and prediction displace dignity, dialogue and professional judgement. The Person-Centred Practice Framework clarifies that humane integration depends on policy, professional formation, organisational culture, governance and the direct care encounter. AI should function as an adjunct to relational care, not a substitute for it. Its value rests on whether healthcare systems use it to amplify human agency, ethical discernment and personhood.
Source: npj digital medicine
Image Credit: iStock
References:
Fernandes JB, McCormack B (2026) Artificial intelligence and person-centred practice: a critical reflection. npj Digit. Med. doi: 10.1038/s41746-026-02776-2
