By Default
Publication Date: 2026-04-05 09:00:00
Editor’s note: Artificial intelligence is increasingly used in the national security community, with applications as diverse as logistics and targeting. However, MIT’s Joel Brenner claims that AI cannot solve the problem of intelligence surprises, arguing that data limitations and the nature of AI thinking will inevitably lead to gaps.
Daniel Byman
***
The amazing ability of artificial intelligence (AI) to do credible human work and radically improve business management and military tactics risks creating a seductive belief in its ability to single-handedly surpass human judgment in strategic decision-making. This is a dangerous illusion. The more consequential a decision is, the more people don’t want machines to make it alone. As I note in a current piece in international security, This is particularly true when it comes to avoiding intelligence surprises.
AI is capable of perfect recall and producing excellent summaries of the available information.
