AI + TCS Working Group

Released September 28, 2026, "AI and TCS: The Next Six Months" is the product of a two-day working group held September 9–10, 2026 at the Simons Institute. The participants considered four questions around how the TCS community can adapt in light of ongoing advances in AI. Via a series of facilitated small and large group discussions, they arrived at the twelve concrete near-term recommended actions presented here, which received broad consensus from the group. The questions as well as the discussions were informed by 134 responses to a survey circulated broadly by the Simons Institute in advance of the meeting.

1. Standardized AI methodology section. Every publicly posted research paper should include a standardized AI Methodology section describing what the authors and AI tools contributed. View template.

2. Tracing the lineage of AI-generated ideas. Authors should make their best effort to trace the lineage of AI-generated ideas, and explain their connections to prior work.

3. Conference rules and norms. Calls for papers should contain clear rules and guidelines, including the following:

   1. AI use is permitted. AI use in the ideas or writing of the submission is permitted, and reviewers may not penalize a paper for AI use as long as it does not inhibit human understanding;

   2. Author attestation. Submission platforms should require an attestation from the authors certifying that they understand and have verified the work they are submitting, and accept responsibility for the correctness and presentation of the work;

   3. Requirement for an overview of results and key technical ideas. Submissions must contain clear exposition, motivation of work and ideas, lineage of ideas, and effective presentation of the key ideas at the beginning of the paper, in the first 10–12 pages.

   4. Poor exposition is sufficient grounds for rejection. Papers may be rejected with a one-sentence review, regardless of the technical merit of the paper, solely because of exposition that is sufficiently unclear.

   5. Submission limit. Impose a limit (recommendation: 5) on how many submissions an author can make to a conference.

   6. AI assistance in reviewing, with disclosure. Reviewers may use AI assistance to help check technical claims (proofs, references, consistency) but the assessment of novelty, significance, and interest must be the reviewer’s own.

4. arXiv posting with submission (experimental). Conferences should require that conference submissions be published on the arXiv.

5. Video talks with submissions (experimental). Conference submissions should require the submission of a talk video in addition to the paper.

6. Unified public repository for conference rules and norms. Establish a coordinated process whereby conferences establish standardized CfPs, along with an up-to-date public repository for publicizing this work.

7. Best works and technical talks in interviews.

   1. Emphasize quality over quantity. Evaluate candidates on no more than three  works of their choosing, with this fact explicit in the instructions for candidates and their letter writers.

   2. Test technical depth. Evaluate the depth of the candidates’ understanding of the aforementioned works more than was done normally until now.

8. Additional postdocs. Establish additional postdoctoral positions for the TCS community, a temporary measure that will be phased out after a few years.

9. Workshops for identifying new theory directions. First, organize a workshop (or several) with the goal of identifying a list of modern topics related to AI; and second, run a series of workshops on each of these topics to identify fundamental directions and establish foundations and agendas for future work.

10. “TCS in the Age of AI” series at the Simons Institute. Establish a recurring and widely accessible “TCS in the Age of AI” series at the Simons Institute.

11. Opportunities for training and evaluation without AI assistance. To ensure development of necessary skills for a successful career in academia, a PhD in TCS should partially include training and evaluation components where the use of AI is forbidden.

12. Coverage of reasonable LLM expenses from research grants. Relevant professional organizations should investigate the barriers researchers have faced in obtaining access to the state-of-the-art AI tools.

Forthcoming