Perhaps the easiest path to making a nuclear weapon, for a country or terrorist seeking one, is to extract a sufficient amount of weapons-grade, highly enriched uranium (HEU) from the nominally peaceful fuel in a research reactor, the small type operating in dozens of countries, including many that lack larger nuclear power plants. According to the late Manhattan Project physicist Luis Alvarez, even high school students “would have a good chance of setting off a high-yield explosion simply by dropping one half of the material onto the other half.” That is why the U.S. nearly half a century ago initiated a program to gradually eliminate such dangerous fuel from these facilities. Now, however, in a stunning reversal, the U.S. Energy Department is actually increasing the likelihood of that deadly scenario by supplying a new research reactor with enough weapons-grade uranium for a sizable nuclear arsenal.

The danger is not just hypothetical. In 1990, Iraqi president Saddam Hussein secretly ordered a crash program to extract HEU from his foreign-supplied research reactor fuel to make an atomic bomb—after his invasion of neighboring Kuwait—but a U.N. intervention fortunately evicted his troops and interrupted the plot before it could succeed.

To prevent such grave risks, the U.S. government since the 1970s has spearheaded an international collaboration to eliminate HEU from research reactors by substituting low enriched uranium (LEU) fuel, the type used in nuclear power plants that is unsuitable for nuclear weapons. (LEU is enriched below 20 percent in the chain-reacting isotope uranium-235, making it unsuitable for nuclear weapons, whereas HEU fuel in research reactors typically is enriched to 93 percent, the same as in U.S. nuclear weapons.) The U.S.-led program has helped contain nuclear proliferation and prevent nuclear terrorism by converting 71 reactors in the U.S. and abroad from HEU to LEU fuel, even tiny ones containing only one kilogram of HEU. The U.S. has not built an HEU-fueled civilian reactor since the 1970s, and no other country has done so since the 1990s.

However the Biden administration intends to violate this nonproliferation policy by supplying over 600 kilograms of weapons-grade uranium—enough for dozens of nuclear weapons—to a privately owned experimental research reactor that would be largely funded by the U.S. government. If the project proceeds, other countries will insist on violating the policy too, refusing to accept a double standard. Whether they import HEU from the United States, purchase it from Russia or build their own enrichment plants, the risks of nuclear proliferation and terrorism will grow again.

The U.S. government is providing $90 million of the $113 million cost to build the Molten Chloride Reactor Experiment (MCRE), which aims to research the potential for a commercial version known as the Molten Chloride Fast Reactor. Although no such power plants exist, they would in theory employ a loop of liquid fuel—uranium dissolved in hot salt—to both sustain the fission reaction and transport the resulting heat. Advocates claim that using liquid fuel, instead of the solid fuel now used in all nuclear power plants, would be a more efficient way to produce electricity and heat for industrial uses. This is not an entirely new concept. In the 1960s, a similar Molten Salt Reactor Experiment was tried but largely failed at Oak Ridge National Laboratory—partly in consequence of the corrosive combination of salt, high temperature and radiation—and it left a particularly nasty radioactive waste problem that still persists. Six decades later, the Energy Department has decided to throw good money after bad.

The technical tweak of the MCRE is to utilize “fast” (high-energy) neutrons rather than the “thermal” (lower-energy) neutrons used in all U.S. nuclear power plants and the 1960s experiment. Fast neutrons facilitate the fission of some radioactive, human-made elements produced in reactors and so can reduce slightly the long-lived radioactivity of the nuclear waste created. But fast neutrons are much less able to induce fission in uranium-235, which is essential for the chain reaction to power the reactor. So, the fuel needs a larger percentage of this isotope, entailing higher uranium-235 enrichment than the 4 percent enriched LEU typically used in nuclear power plants.

However, molten salt fast reactors such as the proposed MCRE do not require HEU. This fact is undisputed because both the Biden administration and its private partners acknowledge that a commercial version, if ever built, would use LEU fuel.

So, if the reactor could use LEU fuel, why is the Biden administration funding an HEU version that would violate U.S. nonproliferation policy?

The simple answer is that the administration has prioritized cost over national security. Energy Department officials conceded in a recent correspondence that using LEU fuel for the MCRE would be “fully consistent” with U.S. nonproliferation policy, which is “to refrain from the use of weapons-usable nuclear material in new civil reactors or for other civil purposes unless that use supports vital U.S. national purposes.” Despite this, the Biden administration decided to use HEU “to keep the size of the experimental reactor small” and to reduce the radioactive waste.

The irony is that other countries have voiced identical arguments to lobby for their own use of HEU, but the U.S. government for half a century has rejected such pleas, emphasizing that nonproliferation is worth the extra expense and that the U.S. practices what it preaches. This longstanding U.S. policy of avoiding a double standard has been crucial to garnering foreign cooperation. Sadly, the Biden administration now is switching to a policy of “Do as I say, not as I do”—which is almost certain to fail.

Earlier this year, U.S. experts—including three former commissioners of the Nuclear Regulatory Commission and three former assistant secretaries of state for nonproliferation—warned Energy Department officials that their plan “would undermine the longstanding U.S. policy of HEU minimization, and thereby increase risks of nuclear proliferation and nuclear terrorism,” urging them instead to “suspend further work on the MCRE until your department’s Nuclear Energy office develops an alternative LEU design.”

The last time that shortsighted U.S. officials planned to build an HEU-fueled research reactor, in the early 1990s, “opposition to the use of highly-enriched uranium in the reactor's core led to its cancellation” by President Bill Clinton. The only question is whether Joe Biden will again demonstrate such U.S. leadership, or gratuitously undermine one of the world’s most successful nuclear nonproliferation programs.

This is an opinion and analysis article, and the views expressed by the author or authors are not necessarily those of Scientific American.