Contractual Deepfakes: Can Large Language Models Generate Contracts?

TL;DR

This paper critically examines the limitations of large language models in generating legally sound contracts, emphasizing that their superficial text generation cannot replace nuanced legal reasoning and transaction-specific understanding.

Contribution

It highlights the gap between LLMs' text generation capabilities and the complex reasoning required for drafting valid, enforceable contracts, challenging optimistic claims about their legal utility.

Findings

LLMs produce superficial and potentially inconsistent contracts.

Generated contracts may be enforceable but often lack transaction-specific appropriateness.

LLMs cannot replace legal reasoning and understanding in contract drafting.

Abstract

Notwithstanding their unprecedented ability to generate text, LLMs do not understand the meaning of words, have no sense of context and cannot reason. Their output constitutes an approximation of statistically dominant word patterns. And yet, the drafting of contracts is often presented as a typical legal task that could be facilitated by this technology. This paper seeks to put an end to such unreasonable ideas. Predicting words differs from using language in the circumstances of specific transactions and reconstituting common contractual phrases differs from reasoning about the law. LLMs seem to be able to generate generic and superficially plausible contractual documents. In the cold light of day, such documents may turn out to be useless assemblages of inconsistent provisions or contracts that are enforceable but unsuitable for a given transaction. This paper casts a shadow on the simplistic assumption that LLMs threaten the continued viability of the legal industry.