Story
OpenAI Releases National Security Principles for Government and Military AI Use
Summary
OpenAI has published a set of principles outlining its approach to partnerships with government and military entities. The guidelines detail specific restrictions on the use of its AI technology, including prohibitions on mass surveillance and autonomous weapons systems.
OpenAI on Wednesday published a set of national security principles that clarify its approach to working with government partners and the military use of its technology. The announcement comes as the artificial intelligence company expands its collaborations with the U.S. government and allied nations on projects related to cyber defense and biosecurity.
Developed with input from national security expert David Kris, the principles apply to all current and future national security and law enforcement partnerships. The guidelines establish key contractual restrictions, prohibiting the use of OpenAI technology for mass domestic surveillance, to direct autonomous weapons systems, or for high-stakes automated decisions.
OpenAI has recently established several international partnerships under its Daybreak cyber defense program, granting "Trusted Access" to nations including Australia, Canada, Japan, and several European Union members. The company also maintains a partnership with the UK government for cyber initiatives and recently expanded access to its GPT-Rosalind model for select U.S. government and allied partners focused on public health and biodefense missions.
AdThe publication of these principles follows a period of reported tension between the U.S. government and another AI developer, Anthropic. A dispute arose earlier this year after Anthropic reportedly declined to remove certain safety restrictions on its AI models for military applications, which led to a temporary directive for federal agencies to cease using its technology. According to the source, relations have eased in recent weeks.