The US State Department recently commissioned the first-ever assessment of proliferation and security risk from weaponized and misaligned AI. Researchers at Gladstone AI have just completed that assessment, as of late February 2024. Their report includes an analysis of catastrophic AI risks, and a “first-of-its-kind, government-wide Action Plan” for what we can do about them.

This perspective has been a long time coming. Over eight years ago, in 2015, I joined over 2,000 scientists and futurists in signing an open letter against weaponized AI. I won’t get too much into the current state here, as the Gladstone report is a richer perspective, but we need to all consider the obvious question of whether there exists either the will or ability to rein in the current, potentially dangerous trend. More on this from me later. For now, I encourage everyone to at least watch this short video and read the Executive Summary of the report posted on the Gladstone AI site.