Checking if the AI is Safe for Older Adults
Building Confidence in AI for Older Adults
Assess AI systems used by or around older adults for dependency risk, cognitive vulnerability, autonomy, and wellbeing impact.
Risk Dimensions
Key areas where AI systems may pose risks to older adults.
Cognitive vulnerability
Does the system account for cognitive decline, memory challenges, or decision-making under fatigue?
Digital literacy gaps
Is the system accessible to people with limited digital experience or confidence?
Isolation amplification
Does the system reduce meaningful human contact or increase social isolation?
Automation of care decisions
Are consequential care or health decisions being delegated to automated systems without adequate oversight?
Loss of human oversight
Do caregivers, family members, or advocates retain meaningful involvement in system decisions?
Manipulation or nudging
Does the system exploit cognitive vulnerabilities through persuasive design or dark patterns?
Consent clarity
Is informed consent genuinely obtainable, or are terms of use effectively inaccessible?
Accessibility bias
Does the system disadvantage people with sensory, motor, or cognitive accessibility needs?
How It Works
Four stages from scoping to actionable recommendations.
Define boundaries
Establish what the AI system should never do. Identify non-negotiable protections for older adults.
Envision better design
What would a genuinely supportive, autonomy-respecting system look like for older adults?
Gather perspectives
Consult older adults, caregivers, and advocates directly about their lived experience with the system.
Document findings
Compile the audit report with evidence, risk scores, and actionable recommendations.
Ready to improve your product?
Turn audit findings into inclusive, dignity-preserving design improvements with our research and design service.