Co-audit: tools to help humans double-check AI-generated content

TL;DR

This paper discusses the development of co-audit tools that assist users in verifying AI-generated content, especially complex outputs like spreadsheets, emphasizing their importance for ensuring quality and correctness.

Contribution

It introduces the concept of co-audit tools, outlines principles for their design, and highlights research challenges in creating effective verification tools for AI outputs.

Findings

Co-audit tools help users verify complex AI-generated content.

Principles for designing effective co-audit tools are proposed.

Research challenges include ensuring accuracy and usability.

Abstract

Users are increasingly being warned to check AI-generated content for correctness. Still, as LLMs (and other generative models) generate more complex output, such as summaries, tables, or code, it becomes harder for the user to audit or evaluate the output for quality or correctness. Hence, we are seeing the emergence of tool-assisted experiences to help the user double-check a piece of AI-generated content. We refer to these as co-audit tools. Co-audit tools complement prompt engineering techniques: one helps the user construct the input prompt, while the other helps them check the output response. As a specific example, this paper describes recent research on co-audit tools for spreadsheet computations powered by generative models. We explain why co-audit experiences are essential for any application of generative AI where quality is important and errors are consequential (as is common in spreadsheet computations). We propose a preliminary list of principles for co-audit, and outline research challenges.