Keeley Erhardt

Graduate Research Assistant, Massachusetts Institute of Technology

2024 Next Gen National Security Fellow

Keeley Erhardt is a doctoral candidate at the Massachusetts Institute of Technology, focused on AI for large-scale network analysis. Her research uses mathematical models to uncover hidden influence in dynamic networks, particularly in social media and information spaces. Her work has significantly contributed to understanding complex information operations.

Collaborating with MIT Lincoln Labs and the Department of the Air Force through the DAF-MIT AI Accelerator, Keeley's research on covert information operations and their impact on the narrative surrounding Uyghurs in China won a Best Paper award at the 2022 SBP-BRiMS conference. She has also published research on Russian influence campaigns and studied the spread of state messaging related to Russia’s 2022 full-scale invasion of Ukraine. Furthermore, Keeley has developed AI tools for uncovering hidden patterns in financial networks.

Professionally, Keeley was a Director of Engineering at Rebellion Defense, leading AI projects for national security. She previously worked at Improbable, focusing on developing virtual simulations for consumer and national security applications. She has shared her knowledge at industry conferences and workshops, discussing topics such as virtual world creation, disinformation challenges, and the use of machine learning in understanding information spread.

Keeley earned her M.Eng. in Electrical Engineering and Computer Science and a B.S. in Computer Science from MIT. She received the Ripple Fellowship for her master's thesis on privacy in digital systems and the Pillar AI Fellowship for her AI research at the MIT Media Lab. Keeley also contributed to the Defending Digital Democracy Project at Harvard Kennedy School's Belfer Center, which focused on digital threat mitigation to safeguard the U.S. election process.

Keeley aspires to continue innovating in AI to bolster the U.S. national security posture.

Support CNAS