The vulnerability of older adults to financial exploitation has raised serious concerns. Effective prevention strategies must consider the intersection of technology and the psychosocial challenges faced by older adults, making them susceptible to exploitation. Smart assistive technology uses data-driven algorithms and generative AI to provide proactive feedback to users or caregivers about fraud warning signs. Unlike existing fraud-detection technologies focusing on individual instances, such as an email or a social media request, the proposed technology tracks all interactions from the same soliciting party, providing a comprehensive safeguard. This project integrates social psychological factors into smart assistive technology design, enabling researchers to study ergonomics, privacy, loneliness, perceptions and attitudes toward technology, and fraud susceptibility. The project will benefit smart in-home product designers in creating practical solutions to mitigate financial exploitation among older adults. It will also benefit society by improving the quality of life of older adults and enhancing their ability to age in place, thus significantly reducing healthcare and long-term care costs. The proposed project will be conducted in five counties in Kansas, impacting over half a million adults aged 60 and older, with a scalable framework beyond Kansas. This aligns with the NSF’s mission to advance national welfare by protecting a vulnerable population from financial fraud and scams.<br/><br/>By integrating social science research into smart assistive technology, this planning project pushes the boundaries of new knowledge. It aims to understand how generative AI models can be tailored for different user profiles (e.g., married vs. single, urban vs. rural, high vs. low socio-economic status) and incorporate various forms of communication (e.g., emails, social media, financial records, phone calls), social-emotional experiences, and socio-demographic factors to combat financial exploitation. The proposed Large Multimodal Model (LMM) framework will expand knowledge on generative AI and fraud prevention, accommodating images and voices in addition to text. The models will reside on users’ mobile devices to address privacy concerns and preferences. This project will also explore how fraudulent solicitations, social-emotional, and socio-demographic factors are linked to susceptibility to financial exploitation among older adults. By examining these relationships, the research aims to develop more effective and tailored interventions to protect this vulnerable population from financial fraud.<br/><br/>This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.