April 29, 2026
Artificial intelligence is rapidly transforming the intelligence community—but it is not replacing the human element anytime soon. That’s the key message from former CIA case officer and RAND researcher Tom Mulligan, who explains that while AI is becoming more powerful and widespread, human intelligence, or HUMINT, remains essential and may even grow in importance.
Mulligan challenges the common belief that AI is a brand-new concept. He points out that as far back as 1964, the CIA was already tracking Soviet developments in artificial intelligence, including early forms of machine learning and pattern recognition. What makes today different is not the existence of AI, but its maturity and accessibility. Modern tools, especially chatbots, have brought AI into everyday use, accelerating its integration across government, industry, and society.
In this environment, Mulligan outlines five key lessons for intelligence professionals. First, case officers must develop a baseline understanding of AI. They do not need deep technical expertise, but they do need enough familiarity to recognize relevant information and act on it. This awareness helps ensure they can identify valuable insights related to AI developments when they encounter them.
Second, AI introduces new risks around deception. Mulligan emphasizes that AI can generate highly persuasive but completely false information. While fabrication has always been a challenge in intelligence work, AI dramatically lowers the barrier, making it easier and cheaper to produce convincing false narratives. This creates a “brave new world” where distinguishing truth from fiction becomes more difficult.
Third, AI is transforming surveillance. Technologies like cameras and sensors already generate massive amounts of data, but AI enables that data to be analyzed in real time and at scale. This increased surveillance pressure means case officers must refine their tradecraft. Even small, unnatural behaviors can now be detected by AI systems, forcing intelligence professionals to operate with greater precision and discipline.
At the same time, AI can also support intelligence operations. Mulligan explains that it can help validate agents by identifying subtle cues, such as micro-expressions, that may indicate deception. These tools can enhance the ability to assess whether a source is trustworthy, adding a new layer of analytical capability.
Another major advantage of AI is its ability to improve communication and persuasion. Mulligan highlights that AI can tailor messages to individuals based on their preferences, personality, and background. For case officers, whose work often involves building relationships and convincing people to share sensitive information, this capability can be especially valuable.
Despite these advances, Mulligan is clear that human judgment remains irreplaceable. As AI tools become cheaper and more widely available, even adversaries will have access to similar capabilities. This widespread adoption increases the relative value of uniquely human skills, such as intuition, experience, and independent thinking.
Ultimately, Mulligan argues that the future of intelligence is not about humans versus machines, but about how they work together. AI will change how intelligence is gathered and analyzed, but it will not eliminate the need for human insight. Instead, it reinforces the enduring—and possibly growing—importance of human intelligence in an increasingly artificial world.