The Signal
Harvard team gets $12 million for AI-designed quantum error correction
A Harvard-led collaboration received $12 million in federal funds for research combining artificial intelligence with quantum error correction. The team plans to develop codes and decoders for scientific workloads and test them on quantum hardware.

Federal award backs application-focused error correction
A $12 million federal award will fund a Harvard-led effort to use artificial intelligence to tailor quantum error correction to scientific workloads, the Harvard Gazette reported on November 20. The initiative, Application-Aware Error Correction Codesign for Scientific Quantum Computing (ASQC), is one of the Genesis Mission’s Phase II projects.
The U.S. Department of Energy’s October 8 announcement listed 12 Phase II awards totaling $159 million, including ASQC. The project is led by Harvard University and is administered through the department’s Advanced Scientific Computing Research division.
AI models and hardware tests
Mikhail Lukin, co-director of the Harvard Quantum Initiative, is ASQC’s principal investigator. The Harvard team includes physicists Markus Greiner and Susanne Yelin, alongside researchers from MIT, Cornell University, Los Alamos National Laboratory, Lawrence Berkeley National Laboratory, Sandia National Laboratories, QuEra Computing, NVIDIA and Anthropic.
The project will train AI models to identify which errors matter for scientific tasks in quantum simulation, materials science, chemistry and nuclear science. The team plans to use AI to design error-correction codes and decoders for those tasks, then test the results on hardware at Harvard, MIT and QuEra.
The work is aimed at quantum error correction, a method for addressing errors in quantum computers. The Gazette says environmental noise and the instability of qubits pose a challenge for quantum computing. It reports that earlier research by Lukin and collaborators suppressed errors to a point where adding qubits would reduce errors rather than increase them.
Research partnership spans universities, labs and industry
The collaboration includes MIT researchers Vladan Vuletić, Soonwon Choi and Hengyun “Harry” Zhou, and Cornell researcher Eun-Ah Kim. Los Alamos participants are Andrew Sornborger, Lukasz Cincio and Michael McKerns; Lawrence Berkeley participants are Wibe de Jong and Costin Iancu; and Sandia participants are Andrew Landahl, Riley Chien and Kevin Thompson.
QuEra’s named investigators are Nathan Gemelke and Michael Gullans. NVIDIA’s co-investigator is Krysta Svore, and Anthropic’s is Alek Kemeny. At Harvard, the project builds on collaborations spanning universities, national laboratories and companies, with the planned work combining AI-designed error-correction methods and tests on quantum hardware.
The Gazette reported that the new project will develop an AI-driven, fault-tolerant error-correction system that can adapt to scientific problems relevant to the Energy Department. The project’s stated aim is to establish scientific foundations for useful fault-tolerant quantum computers.
Sources
- Original source · Press release Harvard Gazette news.harvard.edu
Other pages read
- U.S. Department of Energy energy.gov
Related stories
- UKRI backs Bonato project for real-time quantum-sensor experiments
- Quantinuum and JPMorganChase report error-corrected algorithm runs
- ETH Zurich team teleports logical states on a 41-qubit circuit
- IBM says million-round repetition code can run in one circuit
- Google award backs effort to verify quantum computations