Letter to the Editor, Pleasanton Weekly
Felicia Ziomek
The Department of Energy’s Genesis Mission will use artificial intelligence to speed up scientific research and “national security” work. The White House compares its urgency and ambition to the Manhattan Project: the program that created the first atomic bomb.
Lawrence Livermore National Laboratory, already a central part of the U.S. nuclear weapons complex, will play a major role.
That should alarm us. LLNL-led projects will use AI in many ways. DOE talks about working faster as though faster always means better. At a nuclear weapons laboratory, it does not.
Safety depends on careful human judgment and time. AI cannot understand the actual conditions inside an aging nuclear facility. It may miss failing equipment, inadequate staffing, poor training, waste-handling problems, or other work happening nearby.
If LLNL uses AI to rush safety reviews or experiments, workers and surrounding communities will bear the risk.
LLNL already operates behind enormous secrecy. AI could make its decisions even harder for workers, regulators, and the public to understand or challenge. It could also allow the Lab to design and modernize nuclear weapons faster while avoiding meaningful public oversight.
This is especially dangerous now that New START, the last major treaty limiting U.S. and Russian strategic nuclear arsenals, has expired. Pope Leo XIV called for a concrete and effective replacement and warned leaders to do everything possible to prevent a new arms race. Genesis must not be allowed to fuel that race.
LLNL and DOE must publicly explain how AI will be used and require independent safety reviews for every high-risk decision. Technology should be used to protect life; not to build nuclear weapons faster.