The U.S. Department of Energy is stumbling in its obligation to build an advanced nuclear reactor, putting the country at risk of falling behind China, Russia and other nations in developing this vital low-carbon power technology, the government’s fiscal watchdog warned this week.

The Government Accountability Office (GAO) noted in a report released on Monday that the DOE risks missing  a deadline to complete a prototype by 2021, as mandated by the Energy Policy Act of 2005.

Advanced reactors depart from the conventional “light water reactors” that have defined the commercial nuclear industry for its 50-plus years. They are highly regarded for their potential to improve safety, reduce costs, to yield much less nuclear waste, to burn “waste” such as plutonium as fuel, and to reduce the threat of weapons proliferation.

In principle many of them operate at much higher temperatures than traditional reactors, making them not only more efficient as electricity generators, but suitable as sources of clean heat to drive industrial processes that today rely on CO2-emitting fossil fuels. They could also help produce hydrogen—as the 2005 act proposed—and desalinate water. The are generally well suited to the small, affordable sizes that are coming into vogue.

The industry refers to them as “next generation nuclear power” (NGNP) and “GenIV” reactors. There are several basic types, including one that uses liquid fuel and others that use solid fuel shaped as pebbles or bricks rather than today’s solid rods. They tend to rely on coolants other than water such as gas, liquid salts, or liquid lead. Some designs call for using thorium fuel instead of uranium, and many deploy a “fast neutron” cycle that cuts way down on waste and can burn it.

But the DOE’s Office of Nuclear Energy (NE) has dithered in its commitment to any of them, according to the report.

“DOE has not selected initial reactor design parameters or reported to Congress on an alternative date for making this selection,” the GAO stated. “Without doing so, it is not clear when NE is going to take this next step in deploying the NGNP prototype reactor and it risks the project not being completed by the targeted date in 2021.”

It adds: “Not deploying a prototype carries certain risks, including waning U.S. influence in the safe operation of nuclear plants internationally and potential loss of certain knowledge and expertise.”

Many people in the U.S. share that concern. When last week’s opening plenary of the American Nuclear Society annual meeting in Reno met to assess the future of U.S. global nuclear influence, “the general consensus was that right now we’re holding our own, but there was some concern that we might be losing some ground if we don’t start working in earnest to advance these concepts in the U.S.,” ANS president Michaela Brady Raap told Fortune.

She said the U.S. needs advanced reactors “to maintain its position and to be innovative.”

The GAO account cites a report from the DOE’s own Nuclear Energy Advisory Committee noting  that “the United States risks falling behind other countries—such as Japan, Russia, China, South Korea, and France—that are actively working to deploy and commercialize advanced reactors.”

China, for instance, is developing several, including a solid-fuel “pebble bed reactor” cooled by liquid salt, and a “molten salt reactor” that is both fueled and cooled by salt fluids. It expects to build small prototypes by around 2019 and has put princeling Jiang Mianheng, the son of former Chinese president Jiang Zemin, in charge. It is providing what is believed to be $400 million for the two reactors.

The GAO pointed out that the DOE stopped developing a high temperature, gas-cooled reactor in 2011 because it was unable to agree to cost sharing with industry after spending over $500 million starting in 2005.

Since then, DOE funding has been minimal.

The DOE entered a collaboration with the Chinese Academy of Sciences in late 2011 for development of salt-cooled high temperature reactors. China’s Jiang is co-chair of that partnership, along with the DOE’s Assistant Secretary for Nuclear Peter Lyons. As part of that cooperation, the DOE gave $7.5 million to three U.S. universities—the Massachusetts Institute of Technology, the University of California Berkeley, and the University of Wisconsin-Madison. It tapped Westinghouse, the U.S. nuclear unit of Japanese conglomerate Toshiba, as a commercial partner.

Some people worry that the DOE is giving away technology to China that the U.S. developed in the 1960s, when Tennessee’s Oak Ridge National Laboratory (ORNL) built an experimental molten salt reactor. (President Nixon scrapped it in favor of nuclear technologies that yielded plutonium, desirable for bombs during the Cold War, and also for use in future nuclear reactors.) ORNL is part of the collaboration.

The 1960s ORNL work underpins the molten salt reactors under development today around the world. In the U.S. that includes projects at startups Flibe Energy in Huntsville, Ala., and Transatomic Power in Cambridge, Mass. Likewise, almost all of the other advanced reactor designs date back decades, but have never won adequate financial backing from the conservative international nuclear giants like Westinghouse and France’s Areva. (GE Hitachi has developed an alternative reactor it calls PRISM, but it has yet to sell one.)

U.S. advanced reactor companies include General Atomics in San Diego and TerraPower, the Bill Gates-chaired startup. Several companies are developing small reactors that conform to more conventional designs and do not meet “GenIV” descriptions. NuScale Power, and Babcock & Wilcox each won DOE backing of over $200 million over the last year-and-a-half.

Pro-nuclear environmentalists say that the absence of advanced nuclear is striking given nuclear’s potential to help wean the world off of fossil fuels.

One problem is that the Nuclear Regulatory Commission’s focus on conventional reactor safety deters innovation. DOE’s Lyons noted in a June 6 letter to the GAO that “the NRC’s lack of engagement in this area has resulted in significant delay in their delivery of several major regulatory documents which were promised to DOE for delivery in 2013.”

Regulators in other countries could take a more adaptive approach including in Canada where startups including Terrestrial Energy, Northern Nuclear, and Thorium Power Canada are developing reactors. Russia is planning a domestic energy future around fast reactors. China could outpace them all.

“The driver here is not the technologies, but the political will,” said ANS’ Brady Raap.

Money would help too. Proposal: Given that the oil industry could use heat from small reactors to help process petroleum, perhaps it could provide the wherewithal.