Mind the Gap Foundation Models and the Covert Proliferation of Military Intelligence, Surveillance, and Targeting
Source
New paper with @sarahbmyers & @mer__edith. We challenge the narrative emphasising AI bioweapons risks, and bring attention to the covert proliferation of military intelligence, surveillance, targeting, and reconnaissance (ISTAR) already occurring via foundation models. 1/5
The singular emphasis on AI bioweapons risks leads to ineffective policy frameworks, such as compute thresholds, and neglects the risks that personal data embedded within commercial AI models poses in the context of ISTAR capabilities. 2/5
AI ISTAR uses are already in deployment and are posing risk to civilian life, and the lack of consideration of personal information in the same vein as chemical or biological data within foundation models neglects a crucial proliferation risk vector requiring urgent action. 3/5
We also show how the present and proposed use of foundation models in military settings expands these risk vectors, and create significant vulnerabilities in the larger defense infrastructures these models interface with, compromising national security. 4/5
We conclude that in order to secure military systems and limit the proliferation of AI-based armaments, it may be necessary to insulate military AI systems and personal data from commercial foundation models. You can find our paper on arxiv here: