10173463HRMNY: an Empathic AI solution to transform Mental Health and Capacity Assessment at scaleClosedFast Start ResponseInnovate UKHRMNY was created to solve one of the most urgent but overlooked problems in UK health and social care: the broken statutory assessment processes. These provide legal safeguards for individuals lacking capacity to consent to care arrangements in a care home or hospital, and are used to determine what support they receive under the Care Act. Right now, the system is failing. Four in five assessments miss legally required 21-day deadlines, with delays averaging 156 days; and devastatingly more than 49,000 people died in 2022-23 while waiting for a decision, with their liberty never protected. Families are left in distress, frontline staff are burnt out, and Local Authorities face spiralling legal and financial risks. Our first step was to create a rules-based digital platform, with a contract with Action First to deliver over 20,000 assessments per year. Engagement with Action First has shown the potential for our technology to replace outdated systems, involving repeated sending of zip files, and address challenges within the current system. But to solve the problem at scale, we need AI. This project develops a new AI-powered version of HRMNY. Using a specialised language model trained only on health and social care law, it will: * Automate repetitive form-filling. * Extract key information from clinical notes and reports. * Flag urgent cases and potential errors. * Provide transparent, explainable recommendations to build trust. * Monitor fairness across different demographic groups. All features are co-designed with assessors, doctors, advocates, people impacted and their families, and regulators, ensuring that the system is practical, safe and acceptable. The benefits are huge. For Local Authorities and the NHS, HRMNY could cut delays from 156 to 21 days, clear the backlog of 123,000 cases, and save millions in staff costs and avoided legal challenges. For frontline staff, it reduces paperwork and stress, giving them back time to spend with people rather than forms. For vulnerable people and their families, it means quicker decisions, protection of their rights, and less time in legal limbo. By using technology responsibly, HRMNY aims to restore trust in the system, make safeguarding fairer, and support the UK's health and social care workforce. This project creates the foundation for a scalable, AI-driven platform that can improve assessments not only across the UK but worldwide.