The Future of Data: Preparing for Agentic AI
April 6, 2026 1:00 PM to 2:00 PM
On Monday, April 6, the INSA Foundation hosted the first installment of the 2026 Future of the IC Workforce: Enhancing Intelligence Operations webinar series, underwritten by Booz Allen and supported by ClearanceJobs.
The kick-off program, The Future of Data: Preparing for Agentic AI, included Linda Weissgold, Former Director of Analysis, CIA, and Director, National Security and Intelligence Program, Texas A&M Bush School DC; Jim Hoskins, IC CTO, Hewlett Packard Enterprise; and moderator Lindy Kyzer, VP of Content, ClearanceJobs.
Recap and recording courtesy of our underwriter, Booz Allen!
The conversation addressed the transition from traditional large language models to agentic AI. Jim Hoskins described these systems as autonomous teams that can execute complex tasks rather than just predicting the next word in a sentence. While standard AI might summarize a document, agentic AI can navigate various data repositories and cross-reference signals. Linda Weissgold noted that she looks at it as a team of agents that can work and talk to each other.
Despite these technical leaps, the panel agreed that AI is an enhancement to human judgment, not a substitute. Weissgold noted that the fundamental nature of intelligence remains the intelligence question, which requires human curiosity and the ability to explain the why behind a conclusion. She emphasized the personal responsibility of the analyst, stating, "Ultimately you as the human are responsible. You are that last line of defense. It would not be fair to policymakers if we come back and we say, 'Oh well we got that out of the black box, I do not know why it told me so.'"
The speakers identified "need to know" protocols and data silos as the primary hurdles for implementation. For autonomous agents to provide value, the IC must address policy barriers regarding data access and standardization. Hoskins suggested that the IC might eventually require AI based access control to manage how agents interact with sensitive information. He observed that "Getting through all those policies is probably the biggest challenge to using agentic AI within the IC."
Special thanks to Booz Allen for underwriting the series!

Thank you to our supporting partner!

