Ethics and Accountability in AI for OT
Navigating ethics and accountability when using AI in occupational therapy (OT) is critical to ensure safe, effective, and legally compliant practice.
Shared Responsibility Between Clinicians and AI Developers
While AI can automate documentation, analyse movement, or generate therapy activities, OTs remain responsible for clinical decisions. AI developers must ensure systems are designed ethically, free from bias, and validated across diverse populations. However, final accountability for patient outcomes rests with clinicians, who must apply professional judgement before relying on AI recommendations.
Ensuring Transparency in Clinical Use
AI tools should never operate as “black boxes.” OTs need to understand how algorithms generate outputs, what data is being processed, and what limitations exist. Transparency allows clinicians to explain AI-informed decisions to patients, schools, or NDIS reviewers, maintaining trust. Where AI is used in documentation, clinicians must review drafts before approval to ensure accuracy.
Privacy and Data Protection Obligations
Because AI systems often process sensitive patient data, compliance with the Australian Privacy Principles (APPs) is mandatory. This includes obtaining informed consent, minimising data collection, encrypting records, and ensuring secure storage. OTs must verify that AI vendors provide role-based access, audit trails, and clear data handling policies. Accountability extends to both the tool provider and the clinician using it.
Ethical Use in Client-Centred Practice
Ethics in OT requires care to remain person-centred. AI should enhance—not replace—the therapeutic relationship. For example, an AI-generated sensory diet or handwriting analysis must be contextualised within the child’s environment, family goals, and clinical reasoning. Clinicians must guard against over-reliance on automation and always prioritise client dignity and participation.
Regulatory and Professional Standards
In Australia, accountability for AI use sits within the frameworks of AHPRA, RACGP, and NDIS Quality and Safeguards Commission. OTs must ensure AI use aligns with professional codes of conduct, clinical governance structures, and funding compliance standards. Failure to apply due diligence may expose practices to ethical breaches or audit risks.
Conclusion
Ethical AI use in OT requires a shared responsibility model—developers must design fair, secure systems, while OTs remain accountable for applying clinical judgement and safeguarding patient trust. In Australia, embedding compliance with APPs and professional standards ensures safe, transparent, and patient-centred outcomes. Therefore, AI should be seen as a supportive tool, not a replacement for professional accountability.