A US State Department-funded study authored by the consultancy firm Gladstone AI says the government should consider a temporary ban on artificial intelligence (AI) that surpasses a certain computational power threshold. The study’s authors warn that advanced AI poses an extinction-level threat to humanity.
Their 247-page report proposes the enactment of sweeping government powers to regulate the development of AI as the technology could “destabilize global security” by hijacking nuclear weapons and critical infrastructure. Gladstone AI suggests the executive branch be granted new emergency powers to respond to hypothetical AI threats.
The State Department commissioned report also recommends treating high-end computer chips crucial to AI development as international contraband and implementing strict monitoring of hardware usage. Gladstone AI’s conclusions echo sentiments expressed by some in the technology industry, government, and academia who warn that while AI holds significant potential, mismanaged deployment could be radically disruptive.
Gladstone AI’s safety report follows recent concerns raised by UNESCO over neurosurveillance and mental privacy infringements relating to emerging brain chip technology. The AI report was prepared for the State Department’s Bureau of International Security and Nonproliferation, tasked with studying and curbing the threat of emergent weapons systems.
Mark Beall, one of the report’s co-authors, has since left Gladstone AI to launch a new Super PAC, Americans for AI Safety. Beall, the former DoD AI strategy chief, and his Super PAC hope to make AI safety “a key issue in the 2024 elections, with a goal of passing AI safety legislation by the end of 2024.”