AI might pervade and transform our lives as much or more than the industrial revolution. This excitement is abated by concerns about the risks associated with AI becoming more capable and autonomous. It is widely accepted that AI possesses a different—perhaps “alien”— type of intelligence. So, it is critical that we understand which AI systems already possess intentions, or may in the future. And we must make sure that those intentions align with human goals—before the AIs act on them. What is more, if we wait until the risks are imminent, it might be too late. We must anticipate and be proactive about the risks.

Our mission is to bring the best minds in the world together to bear on this topic—to drive academic research, development, and public engagement on intentions and AI. As AI evolves, matures, and advances, we must stay ahead of it—in perpetuity. We will develop methods to detect, characterize, and intervene on intentions in artificial agents before they act on those intentions. We will anticipate future developments in AI to remain ahead of the issue. Finally, our work on intentions in AI systems will provide insight and enhance research on intentions in humans and in other biological systems.