Auditing and Controlling AI Agent Actions in Spreadsheets DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; June 03–05, 2018; Woodstock, NY

Sadra Sabouri, Zeinabsadat Saghi, Run Huang, Sujay Maladi, Esmeralda Eufracio, Sumit Gulwani, Souti Chattopadhyay · 2026

Read Auditing and Controlling AI Agent Actions in Spreadsheets DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; June 03–05, 2018; Woodstock, NY by Sadra Sabouri, Zeinabsadat Saghi, Run Huang, Sujay Maladi, Esmeralda Eufracio, Sumit Gulwani, Souti Chattopadhyay (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-20070.txt.

Description

Sabouri and six colleagues present Pista, a spreadsheet AI agent that splits execution into auditable, controllable actions. Agents often finish multi-step knowledge work opaquely, leaving users unable to check assumptions, catch errors early, or redirect the work. In spreadsheets the stakes run high, since each agent decision lands directly in the user's own cells. A formative study with 8 users and a summative comparison with 16 users found active participation improved outcomes, comprehension, and co-ownership, surfacing errors post-hoc review would miss.

Catalog record

CREATORSadra Sabouri, Zeinabsadat Saghi, Run Huang, Sujay Maladi, Esmeralda Eufracio, Sumit Gulwani, Souti Chattopadhyay
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE56849
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.20070

Browse the full library · About the Tower of Babel