This article examines the ways in which the US intelligence community is leveraging the power of artificial intelligence (AI) for national security purposes. Drawing on declassified intelligence records, it contends that this community has been fascinated by AI for decades. This is important to acknowledge because this historical context has shaped contemporary projects and thinking within the community. It has given the United States a first-mover advantage, establishing precedents that other global actors need to comply with, negotiate or resist. The article advances three arguments. One, the community has long recognized that it needs to collaborate with the tech sector on AI. However, these relationships bring certain challenges since the sector is a curious compound of ideologies and interests. Two, while the community was initially attracted to the data processing advantages of AI to help human analysts to overcome “data smog,” today it has broadened its focus to consider how AI can improve all stages of the intelligence cycle. Three, while many voices feverishly herald the transformative potential of AI in the global security environment, we argue instead that US agencies will not be able to exploit the full potential of AI, and thus talk of an intelligence revolution is premature.