David Awad · forthcoming, Capital University Law Review (2026)
A Federal Evidentiary Privilege Rule for Human-AI Communications: Proposing a New Federal Rule of Evidence 503.
This Article argues the Federal Rules of Evidence should be amended to create a new privilege protecting confidential communications between people and AI systems from compelled disclosure. With AI now the first and often only legal, medical, and therapeutic advisor available to millions who cannot afford counsel, leaving those exchanges fully discoverable chills the candor those tools depend on and entrenches a two-tier justice system.
David Awad · Draft, April 2026
Prompts, Weights, and Ordinary Meaning: Toward a Daubert Standard for LLM-Assisted Statutory Interpretation.
As courts begin querying large language models to determine the ordinary meaning of statutory terms, this Article shows that LLM outputs are too sensitive to undisclosed inputs — prompts, temperature, system prompts — and too prone to hallucination to be trusted as currently deployed. It proposes a Daubert-style gatekeeping framework requiring disclosure of all inputs, systematic reformulation testing, and a hearing before such outputs are admitted as evidence of ordinary meaning.