Harvard University shieldHarvard UniversityDepartment of Mathematics CMSACenter of Mathematical
Sciences and Applications
Harvard Mathematicians and AINews and Initiatives

Harvard
Mathematicians and AI:
News and Initiatives

This page highlights how Harvard mathematics faculty are engaging with AI through research, teaching, evaluation, and public dialogue. Their work explores what these tools can offer, where they fall short, and what remains essential to mathematical understanding.

Mathematics, together.Doing Mathematics with AI Agents ↗ · Harvard · October 31, 2026

Two summits.
An ongoing conversation.

Two Harvard-hosted summits brought mathematicians together to address the future of research and doctoral education.

These reports are invitations to further discussion. Feedback information is available on the summit pages.

Research summit October 22–23, 2026

Math Research
in the Age of AI

At CMSA, 32 participants from academia, government, and industry developed recommendations for the future of mathematical research.

Organizers
Dan Freed (Harvard), Larry Guth (MIT), Bryna Kra (Northwestern), Emily Riehl (Johns Hopkins), and Mohammed Abouzaid (Stanford).

Education summit September 17–18, 2026

PhD Math Education
in the Age of AI

A Harvard-led gathering of 24 mathematicians developed recommendations to help departments rethink doctoral training as AI reshapes mathematical practice.

Harvard organizers
Dan Freed, Joe Harris, Mark Kisin, and Melanie Matchett Wood.

Events

Math and AI workshops, seminars, and summits hosted or cosponsored by Harvard Mathematics and CMSA.

Showing 6 of 7 upcoming events

NOV212026
WorkshopUpcoming

Doing Mathematics with AI Agents ↗

Organizers: Melanie Matchett Wood (Harvard), Andrew Sutherland (MIT), Saman Habibi Esfahani (Harvard CMSA), and Michał Szachniewicz (Harvard)

Zimmer Hall, Harvard Science Center · 9 am–5 pm ET

About this event

For Boston-area mathematics postdocs. Cosponsored by Harvard Mathematics and CMSA.

All times are Eastern. Follow each event link for current details and participation requirements.

Faculty initiatives

Harvard faculty are helping guide efforts in evaluation, education, mathematical communication, and the relationship between mathematics and the AI industry.

Stewardship

AGMAI

The Advisory Group on Mathematics and Artificial Intelligence advises AI companies on engagement with mathematics, including the responsible presentation and release of results.

Melanie Matchett Wood
Melanie Matchett Wood Advisory Group Member Photo: Harvard Mathematics
Education · National Academies

Data and Computing in K–12 Education

The 2026 National Academies of Sciences, Engineering, and Medicine report Data and Computing in K–12 Education: Foundational Competencies identifies the knowledge and skills all students need to navigate a world shaped by data, computing, and AI.

Portrait of Laura DeMarco
Laura DeMarco Report committee member · Mathematical Sciences Education Board member
Communication

Research repository

Hexagon

Lauren Williams (Harvard) · Board of Directors

A mathematical research repository created in response to AI’s changing role in the field, with disclosure of AI use and human understanding. Deposits are moderated, rather than peer reviewed.

Joint Mathematics Meetings

Modernizing Math Education

Laura DeMarco (Harvard) · TPSE Invited Address

An invited address on what mathematics education can and should offer at every level, including how teaching can respond to the possibilities (and the risks) of AI.

View the JMM address
AMS · Science policy

White Paper on Artificial Intelligence

Mihnea Popa (Harvard) · White paper contributor; AMS Committee on Science Policy and Council member

Popa helped prepare the AMS Committee on Science Policy’s August 2026 white paper. It argues that mathematics can help explain how AI works, measure its uncertainty, check its reasoning, and identify its limits—work that could make AI more reliable and safer. The paper calls for sustained support of mathematical research and education.

Read the AMS white paper (PDF)
National Academies

–

Organizing Mathematical Knowledge in the Age of AI and Formalization

Melanie Matchett Wood (Harvard) · Panel moderator

A national conversation on how AI and formalization are changing the creation, organization, and validation of mathematical knowledge, and the shared infrastructure needed to support that work.

Meeting details & agenda
Undergraduate education

–

Modernizing Undergraduate Math Summit

Brendan Kelly (Harvard) · Co-organizer

Kelly helped organize this Harvard-hosted summit, convened by TPSE Math. Leaders in mathematics, statistics, higher education, and partner fields discussed how undergraduate math can respond to advances in AI, data, computation, and modeling while strengthening students’ mathematical judgment.

Further perspectives

The Harvard Crimson · Opinion · November 5, 2026

Math Is Still Worth Studying in the Age of AI

Harvard mathematics and physics concentrator Evan B. Tsingos ’27 makes the case for studying mathematics as a source of human understanding, creativity, and rigorous thinking in the age of AI.

The FAS Current · October 19, 2026

Top mathematicians grade AI’s ability to solve toughest problems

Lauren Williams and First Proof bring mathematical expertise to the evaluation of AI research capabilities.

Sanlian Lifeweek · Interview · October 11, 2026

Curtis McMullen and James Maynard on AI’s Claims to Solve Longstanding Math Problems

In a joint interview, Curtis McMullen and James Maynard discuss AI-generated results, concerns about credit and research norms, and why mathematical understanding matters beyond obtaining answers.

TechCrunch · October 8, 2026

OpenAI’s math solutions aren’t meeting the field’s standards yet

Melanie Matchett Wood discusses human understanding of AI-generated mathematical results.

The Harvard Crimson · September 22, 2026

Harvard Math Professor Joins Independent Group Advising AI Companies on Research Standards

Melanie Matchett Wood helps guide an independent group advising AI companies on responsible engagement with mathematical research.

Math and AI declaration · September 11, 2026

A Severe Misalignment of AI in Mathematics

Curtis McMullen is among 28 Fields Medalists who signed a declaration calling for AI to support mathematical understanding and the people who advance it. The declaration warns that a race to solve difficult problems can overshadow understanding results, crediting prior ideas, and developing students.

Quanta Magazine · The Joy of Why · June 25, 2026

What Is the Positive Grassmannian and Why Does It Show Up Everywhere?

Lauren Williams discusses the unexpected connections in her mathematics, then turns to First Proof and the possibilities for AI as a research partner.

The FAS Current · June 17, 2026

First Proof’s second batch of math problems test AI

Lauren Williams and collaborators bring rigorous evaluation to AI-generated mathematics, with human referees gathering at Harvard CMSA to assess the results.

The New York Times · February 7, 2026

These Mathematicians Are Putting A.I. to the Test

Lauren Williams and fellow First Proof mathematicians explain their independent effort to assess AI using problems drawn from unpublished research, and why expert mathematical judgment matters.

Watch & Explore

Selected conversations on mathematics and AI at Harvard CMSA.

Ongoing seminar series

Selected talks, conversations & workshops

First Proof: Introduction & results

Dan Freed & the First Proof editors · June 3 & 10, 2026

Two Harvard CMSA webinars introduce First Proof and present its second-batch evaluation of AI-generated mathematics.

Working with LLMs to do high quality math

Daniel Litt · September 9, 2026

A talk in Harvard CMSA’s AI for the Working Mathematician seminar on using large language models in mathematical research.

How could a Superhuman AI mathematician come about?

Sanjeev Arora · March 4, 2026

In CMSA’s 2026 Ding Shum Lecture, Sanjeev Arora explores AI self-improvement, formal verification, and the possibilities for mathematical reasoning.

Machine Learning in Science Education

Panel discussion · September 30, 2024

A panel hosted by Harvard CMSA explores teaching and learning with AI, including student-centered tutors and classroom activities.

Yann LeCun on the future of AI

Special lectures · 2024 & 2025

In two CMSA special lectures, Yann LeCun discusses world models and approaches to AI systems that can learn, reason, and plan.

Mathematical Foundations of AI

Workshop · October 6–10, 2025

A workshop at Harvard CMSA explores mathematical approaches to understanding modern AI and improving its reliability.

Mathematics and AI at Harvard