Nuclear Nonproliferation, Safeguards, and Enrichment
“South Africa repeatedly has stated its commitment to nuclear nonproliferation since it ended its nuclear weapons program and signed the Nonproliferation Treaty in July 1991,” observes a brief new report from the DNI Open Source Center, which also notes that “South Africa has a history of backing Iran’s ‘peaceful’ nuclear development.” See “South Africa’s Nuclear Nonproliferation Posture Remains Consistent” (pdf), Open Source Center Report, 9 April 2010.
“The implementation of international safeguards at the Rokkasho Reprocessing Plant (RRP) in Japan has been the largest challenge the International Atomic Energy Agency (IAEA) has faced to date,” according to a recent report on the development of nuclear safeguards. See “Lessons Learned in International Safeguards — Implementation of Safeguards at the Rokkasho Reprocessing Plant” (pdf), Oak Ridge National Laboratory, December 2009.
An overview of uranium enrichment programs around the world was prepared last year for the Department of Energy, based exclusively on open source information. See “Profile of World Uranium Enrichment Programs – 2009” (pdf), Oak Ridge National Laboratory, April 2009.
At this inflection point, the choice is not between speed and safety but between ungoverned acceleration and a calculated momentum that allows our strategic AI advantage to be both sustained and secured.
Improved detection could strengthen deterrence, but only if accompanying hazards—automation bias, model hallucinations, exploitable software vulnerabilities, and the risk of eroding assured second‑strike capability—are well managed.
New initiative brings nine experts with federal government experience to work with the FAS and Tech & Society’s Beeck Center for Social Impact + Innovation, the Knight-Georgetown Institute, and the Institute for Technology Law & Policy Wednesday, June 11, 2025—Today Georgetown University’s Tech & Society Initiative and the Federation of American Scientists (FAS) announce two […]
A dedicated and properly resourced national entity is essential for supporting the development of safe, secure, and trustworthy AI to drive widespread adoption, by providing sustained, independent technical assessments and emergency coordination.