Artificial intelligence and mental capacity legislation: Opening Pandora's modem.

Area of Science:

• Legal and Ethical implications of Artificial Intelligence
• Mental Capacity Law and Technology Integration
• Healthcare Decision-Making and Assistive Technologies

Background:

• People with impaired decision-making capacity have the right to access technology, including artificial intelligence (AI).
• Ireland's Assisted Decision-Making (Capacity) Act 2015 incorporates 'assistive technology' into its communication criteria for capacity.
• Barriers to advance care planning include lack of knowledge, fear, and time, which AI may help overcome.

Purpose of the Study:

• To explore the realization of technology access rights for individuals with impaired decision-making capacity within AI and mental capacity legislation.
• To analyze the potential benefits and risks of AI in assisting communication under Ireland's Assisted Decision-Making (Capacity) Act.
• To identify principles for the ethical and effective use of AI in supporting decision-making for vulnerable populations.

Main Methods:

• Analysis of Ireland's Assisted Decision-Making (Capacity) Act, 2015, focusing on 'assistive technology' and 'communication' criteria.
• Exploration of generative AI's capabilities and ethical considerations, referencing UN guidance on AI ethics.
• Examination of AI's potential to address barriers in advance healthcare planning and decision-making.

Main Results:

• Generative AI may help overcome barriers to advance care planning, but poses risks of influencing decision content.
• AI models, when prompted to create advance healthcare directives, may inadvertently set an agenda.
• Purpose-designed AI models, including foundational models, offer potential for education, interaction, and professional updates, optimizing benefits and minimizing risks.

Conclusions:

• AI holds significant potential to support individuals with impaired decision-making capacity, particularly in healthcare directives.
• Ethical frameworks and careful design, such as using purpose-designed AI models, are crucial to mitigate risks associated with AI in this context.
• Open discussion and responsible development are needed to optimize AI's benefits while minimizing risks for people with impaired decision-making capacity.