Slides

[https://docs.google.com/presentation/d/1tsdEfVCLIxGutte9-ekCzmmoFqF1Etvl/edit?usp=sharing&ouid=100090609942785485941&rtpof=true&sd=true](https://docs.google.com/presentation/d/1tsdEfVCLIxGutte9-ekCzmmoFqF1Etvl/preview?usp=sharing&ouid=100090609942785485941&rtpof=true&sd=true)

Talk Outline

Though voice AI systems have not yet achieved the technical capacity of text-based LLMs, they are increasingly gaining popularity amongst legal researchers and practitioners. In this talk, we will discuss ways that these systems may accelerate legal workflows and expand access to justice when implemented for legal intake purposes, particularly for individuals who face barriers to traditional legal services. We will additionally discuss experiential discoveries of limitations of these systems, and directions for future fundamental research. We also surface the ethical considerations practitioners must navigate, including informed consent and data privacy.

Speaker Bio

Riya is a recent graduate of Stanford University, leaving with a degree in computer science and mathematics. While at Stanford, she conducted research on AI evaluations for legal workflows, and technology design ethics, particularly around deepfake sexual violence. Her work has been presented and published in top legal and AI research venues, including NeurIPS and ICAIL. Her current research focuses on applying voice AI to access to justice issues.