The National Science Foundation has confirmed its plan for “Grand Research Challenges” and additional research opportunities focused on NSF priority areas, rather than investigator-driven ones. In an open letter last week, acting Director Brian Stone announced the change as part of NSF’s new “portfolio-based approach” to grantmaking. NSF will also announce a series of Grand Research Challenges over the coming months to “take thesis-driven bets that stand to unlock new frontiers of S&T.” Nature reported last month that NSF was allotting as much as $1 billion of its research budget to the challenges, contributing to the issuance of far fewer grants this year.
The letter also says NSF will experiment with novel proposal review mechanisms, such as “golden tickets,” which allow one reviewer to advance a proposal that a larger panel may not approve, and distributed review, in which all applicants to a solicitation serve as reviewers for that solicitation. The agency will also fund more grants with time horizons longer than five years, as well as shorter-duration ones that will have “applications of only a few pages and rapid funding decisions.”
In addition, NSF will establish an Office of Metascience to develop and evaluate new approaches to grantmaking and publicly report the effectiveness of different funding mechanisms. NSF intends for the office to serve as a model for other agencies, the letter says. Through its metascience efforts, NSF aims to “double the scientific productivity and impact generated by each federal research dollar by 2036,” echoing the goal of the Department of Energy’s Genesis Mission on AI. NSF’s budget request for fiscal year 2027 proposes a metascience office but does not specify a funding amount. The only congressional budget proposal for NSF so far, approved by the House Appropriations Committee in May, does not mention a metascience office.
Universities respond to DOD audit demands
At least one university has ended a study abroad program in China following August security audits of 30 schools ordered by the Department of Defense.
In July, Defense officials added three prominent Chinese research universities — Fudan University, Shandong University, and Shanghai Jiao Tong University — to the Section 1286 list of entities “flagged as a concern to national security.” That move barred those institutions from fundamental research collaboration with US schools and led to the August audits.
In response, the University of North Carolina has canceled its study abroad program with Fudan University. Other schools reported they would continue with undergraduate-exchange programs that don’t involve research.
Adding research universities to a list previously dominated by institutions more directly connected to military-industrial industries widened Section 1286’s scope, noted Tech Times. Educational organizations condemned the move as undermining international scholarship and scientific progress.
House Science Republicans send letter to DOE, NIH, and NSF asking about biotechnology
Republican lawmakers last Wednesday asked three federal research agencies for their biotechnology risk-mitigation policies, particularly regarding dual-use technologies, AI, and foreign actors. The letters cite the 2025 National Security Commission on Emerging Biotechnology (NSCEB) final report’s biosecurity and innovation recommendations, asking for agency updates on their implementation. The House Science Committee’s investigations subcommittee will hold a related biosecurity hearing on Wednesday.
The Energy Department, National Institutes of Health, and National Science Foundation separately received letters, each requesting information on their biosecurity monitoring mechanisms, steps taken to address the “risks arising from the convergence of biotechnology and AI,” and more. Biotechnology is moving beyond its role in healthcare, the lawmakers suggested, citing a 2025 Stanford University report describing its emergence in energy, agriculture, and manufacturing.
Senators turn spotlight on OpenAI amid controversy
Sen. Bernie Sanders (I-VT) has scheduled a Wednesday Senate briefing on the dangers of AI, following industry calls for a development slowdown and national safety requirements, amid heightened attention from lawmakers. Last week, Sen. Josh Hawley (R-MO) announced that he is investigating OpenAI’s hack of Hugging Face, an AI data and model hosting company. And Sen. Chris Van Hollen (D-MD) on Thursday sent a letter to OpenAI’s CEO expressing “serious concern” about the safety of its latest models. The consideration from Congress follows wider questions about the company’s research and training practices.
OpenAI has also become embroiled in a research integrity controversy following its claim last week that it used its products to find a proof for one of the exceptionally difficult Millennium Prize Problems in mathematics. Tristan Buckmaster, an NYU mathematician working on the same problem, said OpenAI tried to pressure him to sign onto the proof and cut out a collaborator employed by rival software developer Anthropic. He also said that he and his colleagues stored their work on the problem in Codex, another OpenAI product, intimating that OpenAI’s proof may have been developed using their material, which OpenAI denies.
Also on our radar
DHS is proposing scrapping the 60-day discretionary grace period that allows certain visa holders, including H-1B holders, to remain in the US after leaving or being fired from a job that qualified them for their visa.
The final rule ending “duration of status” for student visas is slated to take effect on Tuesday, though a Massachusetts judge could still pause or strike it down.
The Union of Concerned Scientists and the Environmental Defense Fund sued NSF and OMB last week for failing to provide requested records on the decision to restructure NCAR.
AGU said it will help distribute the Arctic Report Card and support its authors after NOAA pulled out of the annual climate assessment. The report card is slated for release on Dec. 7 at AGU’s annual meeting, as in past years.
Sens. Dave McCormick (R-PA) and John Hickenlooper (D-CO) launched a Science and Innovation Caucus last week that aims to “work together with industry leaders and the research community to develop policies that promote U.S. leadership in science.”
NASA launched a “foundation” AI model for lunar science to help researchers analyze large quantities of data related to the Moon’s surface.
From Physics Today: An interview with a prominent member of the Marshallese diaspora delves into the ongoing impact of aboveground US nuclear weapons testing in the Pacific archipelago.
Know of an opportunity for scientists to engage in science policy? Email us at fyi@aip.org.
Do you have a story to tell about how science policy is impacting you?
AIP’s research team is gathering first-hand accounts from scientists, engineers, students, and staff whose careers have been affected by policy and funding changes over the past year. Volunteers can submit their stories via this online form. Participants’ stories will be added to the Niels Bohr Library & Archives digital repository as searchable, citable records — with options for anonymity and a five-year embargo period. Read more about the initiative here.
Around the Web
News and views currently in circulation. Links do not imply endorsement.