Prestigious NSF graduate fellowship tilts toward AI and quantum

From ScienceMag:

Roughly one in five recipients of the National Science Foundation’s (NSF’s) prestigious Graduate Research Fellowships Program (GRFP) typically works in the life sciences. However, there’s nary an ecologist, neuroscientist, or molecular biologist in the latest batch of 500 fellows announced earlier this month. At the same time, computer scientists captured a much larger share of awards, which provide an annual stipend of $37,000 for 3 years, than they have done in previous rounds.

NSF isn’t saying why this cohort looks so different by field of study than the 1000 fellows announced 2 months earlier. But the new distribution appears to align with President Donald Trump’s push to prioritize artificial intelligence and quantum information science. Some scientists worry it may also reflect Trump’s opposition to any federal investments that broaden participation in science, as most of the students in the latest cohort work in fields that are disproportionately male, white, and Asian.

The program is open to students in any field that NSF funds and has supported 50 Nobel laureates since its debut in 1952. But many scientists saw the size of the initial 2025 class—at less than half previous annual levels—as a retreat from NSF’s historical mission to train the next generation of U.S. scientists and engineers. Their fears grew last month when NSF’s oversight body spoke of its desire to see the private sector sponsor a significant number of future fellows.

NSF’s 13 June announcement of a second round—selected from 3000 applicants awarded “honorable mentions” in the first round—was seen as evidence that the agency hadn’t abandoned the program. However, the striking difference by discipline between the new cohort and the first round (see graphic, below), as well as from previous annual classes, has revived concerns about the program’s future.

An analysis by two former GRFP directors has found that life scientists, for example, were shut out in the second round. Some 214 students in those fields were named fellows in April, and another 962 were designated as honorable mentions, according to data shared with Science by Giselle Muller-Parker and Susan Brennan. Not one of those 962 runners-up was selected for an award in the second round.

Similarly, only seven of the 187 honorable mentions in psychology were named fellows this month, as were only eight of the 114 honorable mentions in the social sciences. Students in the geosciences also fared poorly in the second round, with only 25 of the 242 honorable mentions winning a fellowship.

In contrast, well over half of the 203 honorable mentions in computer science—125—were named fellows this month. Some 101 of the 260 honorable mention students in physics and astronomy won the prize. And 123 of the 594 engineering students initially named as honorable mentions got scholarships.

Congress has tweaked NSF’s mandate for science, technology, engineering, and math (STEM) workforce training over the decades, most recently in 2022 when it told the agency to address demand “in critical STEM fields.” But the definition of those fields encompasses every area NSF has traditionally supported, from designing more advanced semiconductor chips to understanding how humans respond to climate change. Muller-Parker and Brennan say they don’t understand why NSF would suddenly change direction without a clear legislative mandate.

“In the absence of evidence for why one STEM discipline should be prioritized over another … we find this trend to be ill-considered,” Muller-Parker and Brennan write in a joint email to Science. Continuing in that direction, they add, poses “substantial risk” for training the U.S. STEM workforce.

Former fellow Alexandra Harmon-Threatt, a pollinator ecologist at the University of Illinois Urbana-Champaign, agrees. “Not investing across the board is extremely shortsighted and undermines America’s scientific future,” she says. “These students won’t be making major contributions for 5, 10, or 15 years, and it’s hard to know where we will have major gaps that need to be filled at that time.”

An NSF spokesperson declined to comment on its criteria for selection or give a reason for the skewed distribution.

The demographic differences between the fields being favored and shunned add to the challenges already facing students from minority groups, says former fellow Christian Cazares, a neuroscientist at the University of California San Diego. “Students like to think the process is merit based,” says Cazares, who was born in Mexico and was the first in his family to attend college. “And GRFP is one of the most important fellowships for early-career scientists. If it’s now simply another way to carry out the administration’s priorities, that sends a message to these students that maybe they aren’t welcome in science.”

Kelsey Lucas, an aquatic biologist at the University of Calgary and former fellow, sees the skewed distribution as a sign that NSF is emphasizing more applied research. “Science is about pushing into the unknown to tackle big problems,” she says. “By cutting GRFP and steering awards toward a privileged few, we are leaving behind perspectives and ways of knowing that could lead to innovative solutions to the problems we face today.”

Read More