Computer Theorists Draft First Unified AI Usage Rules

The Simons Institute (UC Berkeley) published the working group report “AI and TCS: The Next Six Months” with 12 recommendations on how computer theorists should disclose AI contributions, format submissions, and adapt PhD training for the LLM era.

Compiled by Sergey KostenchukPublished 2026-10-01Updated 2026-10-01

2026-10-01 Research

🤖 Computer theorists given AI usage rules

The Simons Institute (UC Berkeley) published the working group report “AI and TCS: The Next Six Months” following a September 9–10 workshop: 32 participants, 134 survey responses, 12 recommendations. Key points: an AI Methodology section with 6 categories of AI contribution in each paper, up to 5 submissions per author, author attestation, and permitted AI assistance for reviewers with disclosure.

🌍 STOC 2027 — the top conference in theoretical computer science — has already decided to adopt almost all recommendations: AI contribution disclosure rules and submission limits will enter peer review in the coming months.

👤 The report, LaTeX template for the AI contribution section, and action tracking are available for free on the institute’s website. Authors and reviewers should prepare for attestation, and reasonable LLM expenses may be covered by grants.

Source 1: https://simons.berkeley.edu/AITCSreport Source 2: https://simons.berkeley.edu/news/ai-tcs-next-six-months

Sources