When AI output tips to bad but nobody notices: Legal implications of AI's mistakes

TL;DR

This paper reveals that generative AI can reliably produce fabricated legal information due to deterministic internal states, posing risks to legal practice and suggesting the need for verification protocols.

Contribution

It uncovers the deterministic failure mechanism in AI's legal output and proposes verification protocols to mitigate fabrication risks in legal applications.

Findings

AI fabrications are linked to internal state thresholds.

Fabrication risk is a foreseeable consequence, not random.

Verification protocols can reduce legal misinformation.

Abstract

The adoption of generative AI across commercial and legal professions offers dramatic efficiency gains -- yet for law in particular, it introduces a perilous failure mode in which the AI fabricates fictitious case law, statutes, and judicial holdings that appear entirely authentic. Attorneys who unknowingly file such fabrications face professional sanctions, malpractice exposure, and reputational harm, while courts confront a novel threat to the integrity of the adversarial process. This failure mode is commonly dismissed as random `hallucination', but recent physics-based analysis of the Transformer's core mechanism reveals a deterministic component: the AI's internal state can cross a calculable threshold, causing its output to flip from reliable legal reasoning to authoritative-sounding fabrication. Here we present this science in a legal-industry setting, walking through a simulated brief-drafting scenario. Our analysis suggests that fabrication risk is not an anomalous glitch but a foreseeable consequence of the technology's design, with direct implications for the evolving duty of technological competence. We propose that legal professionals, courts, and regulators replace the outdated `black box' mental model with verification protocols based on how these systems actually fail.