US military personnel in San Diego at the US-Mexico border.Credit: Carlos Moreno/NurPhoto/Getty
What are the implications of allowing artificial intelligence (AI) to make critical decisions about life and death in combat? That’s a question that Nicholas Evans, a social scientist at the University of Massachusetts Lowell, was hoping his research could answer — until funding for his grants was cut by the United States Department of Defense (DoD) this month.
The grants were among 91 social-science studies terminated by the DoD, including many that were part of the flagship Minerva Research Initiative, which supports basic social-science research so as to better understand emerging threats to national security.
“One of the brilliant parts” of Minerva is that it takes “the notion of security broadly,” says Leonardo Villalón, a political scientist who studies the Sahel region in Africa at the University of Florida in Gainesville. Minerva grants fund research on global dynamics such as violence, instability, natural catastrophes, human displacement and migration, he says.
The defence department stated in a press release that it was “scrapping its social science research portfolio as part of a broader effort to ensure fiscal responsibility and prioritize mission-critical activities”. Termination notices, seen by Nature, state that the grants no longer served DoD’s “program goals or agency priorities”.
“The big challenge”, says Evans, “is that there is almost nowhere else in the United States where you can get two and a half million dollars to do social-sciences research, and that limits our ability to get funded.” He and his collaborators received US$5.3 million in research grants in 2021 and 2024, as part of Minerva. With the funding cut, he will lose US$4.3 million.
National interest
The Minerva initiative was launched in 2008, and grants are managed by research offices run by the army, air force and navy. A portion of the funds go towards educating students at US military schools and academies in key areas of the social sciences, and many of those grants have also been terminated.
Neil Johnson, a physicist at George Washington University in Washington DC, received termination notices for two grants, each worth about $2.5 million. One of them, close to the end of its five-year term, supported research on how threats, hate and extremism spread through online and offline social networks. The other focused on security threats along national borders.
“The rationale was really weird,” says Johnson. For years, he has participated in calls and briefs at DoD agencies. Among other things, he has advised intelligence officers at military bases of his research findings, from the weaponization of health to gun violence. Now that all stops, he says.
Spending money on military preparedness — on armaments and technology, for example — but not on understanding the nature and causes of potential military conflicts is incredibly short-sighted, says Kathy Baylis, a development economist at the University of California, Santa Barbara. “It kind of boggles the mind,” she says. The Minerva Research Initiative accounts for a fraction of the DoD’s budget. In its 2024 budget request, the department requested $64.3 million for Minerva out of a total budget of $842 billion.
Baylis has also had her Minerva grant terminated. Awarded in 2023, it enabled her to study the effects of climate shocks on food security in sub-Saharan Africa. It was initially guaranteed for three years, with an option of two more. Between the Minerva losses and cuts to grants from the US Agency for International Development, Baylis has lost roughly US$5 million over the past few weeks. Since then, she has been scrounging for money to pay salaries and working out ways to share the limited data that she and her team managed to collect. “They just wasted a whole pile of money that had been spent on research that can no longer be fulfilled,” she says.
Villalón, who was studying the impact of climate hazards on societies in the Sahel, and how those communities were responding to changes, had already spent most of the $1.6 million awarded as a three-year grant in 2022. He and his team had only about $200,000 left over, which would have been used to support data analysis and publication.
