42 1 Hegseth, (2025, March 6). Directing modern software acquisition to maximize lethality [Memorandum]. U.S. Department of Defense. https://media.defense.gov/2025/ Mar/07/2003662943/-1/-1/1/DIRECTING-MODERN- SOFTWARE-ACQUISITION-TO-MAXIMIZE-LETHALITY.PDF 2 J.P. Morgan Asset Management. (n.d.). AI investment: Separating hype from opportunity. J.P. Morgan. https:// am.jpmorgan.com/se/en/asset-management/per/insights/ market-insights/investment-outlook/ai-investment/ 3 Ding, (2022, August 25). Government venture capital and AI development in China. Stanford Center on China’s Economy and Institutions. https://sccei.fsi.stanford.edu/china-briefs/ government-venture-capital-and-ai-development-china 4 Ibid 5 Lunden, (2024, December 27). Why Deepseek’s new AI model thinks it’s ChatGPT. TechCrunch. https://techcrunch. com/2024/12/27/why-deepseeks-new-ai-model-thinks-its- chatgpt/ 6 Chapman, (2025, January 30). Pentagon workers used Deepseek’s chatbot for days before it was blocked. Bloomberg. https://www.bloomberg.com/news/ articles/2025-01-30/pentagon-workers-used-deepseek-s- chatbot-for-days-before-block 7 Defense Manpower Data Center. (2024, December). DOD workforce report: December 2024 [Data file]. U.S. Department of Defense. https://dwp.dmdc.osd.mil/dwp/ api/download?fileName=DMDC_Workforce_Report_ December_2024.xlsx 8 U.S. Department of Defense. (2024). Fiscal Year 2025 Budget Request: Financial Summary Tables (p. 1). https://comptroller.defense.gov/Portals/45/ Documents/defbudget/FY2025/FY2025_Financial_ Summary_Tables.pdf 9 Fortune. (2024). Fortune 500: The largest companies in the U.S. by revenue. https://fortune.com/ranking/fortune500/ 10 See Appendix for methodology & calculations. 11 Sankar, (2024, October 31). The Defense Reformation [Figure 1]. https://www.18theses.com/ 12 Schultz, (2021). Please change the acquisition culture! Defense Acquisition Magazine. https://www.dau.edu/ library/damag/march-april2021/please-change-acquisition- culture21 13 Wright, (2015, January 26). The rise and fall of the unipolar concert. Brookings. https://www.brookings.edu/articles/ the-rise-and-fall-of-the-unipolar-concert/ 14 U.S. Department of Defense. (2022). 2022 National Defense Strategy, Nuclear Posture Review, and Missile Defense Review (p. 24). https://media.defense.gov/2022/ Oct/27/2003103845/-1/-1/1/2022-NATIONAL-DEFENSE- STRATEGY-NPR-MDR.pdf 15 Chief Digital and Artificial Intelligence Office. (n.d.). Our Mission. https://www.ai.mil/About/Organization/#our- mission 16 Brose, (2020). The kill chain: Defending America in the future of high-tech warfare (p. 202). Hachette Books. 17 Walsh and Huber, (2023, October 30). A symphony of capabilities: How the Joint Warfighting Concept guides Service force design and development. National Defense University Press. https://ndupress.ndu.edu/Media/News/ News-Article-View/Article/3568312/a-symphony-of- capabilities-how-the-joint-warfighting-concept-guides- service-for/ 18 Kumar, (2024, December 12). In J. Clark, DOD innovation official discusses progress on Replicator. U.S. Department of Defense. https://www.defense.gov/News/News-Stories/ Article/Article/3999474/DOD-innovation-official-discusses- progress-on-replicator/ 19 Clark, (2024, December 12). DOD innovation official discusses progress on Replicator. U.S. Department of Defense. https://www.defense.gov/News/News-Stories/ Article/Article/3999474/DOD-innovation-official-discusses- progress-on-replicator/ 20 Robertson and Albon, (2024, May 23). Replicator drones already being delivered, Pentagon says. Defense News. https://www.defensenews.com/pentagon/2024/05/23/ replicator-drones-already-being-delivered-pentagon-says/ 21 By ‘leading’ we refer to Gen-AI tools such as LLMs that are trained on mass-data sets and capable of serving a range of applications with a high degree of accuracy. These models, such as ChatGPT, Gemini, Llama, and others differ from smaller models requiring less data and investment dollars to build and scale, but offer tradeoffs in versatility, reliability, and functionality. 22 Hays, (2024, October 14). The U.S. defense and homeland security departments have paid $700 million for AI projects since ChatGPT’s launch. Fortune. https://fortune. com/2024/10/14/us-DOD-dhs-700-million-ai-projects-past- two-years-increase-since-chatgpt-launch/ 23 Gruss, (2019, April 15). Could artificial intelligence save the Pentagon $15 billion a year? C4ISRNET. https://www. c4isrnet.com/it-networks/2019/04/15/could-artificial- intelligence-could-save-the-pentagon-15-billion-a-year/ 24 Bommasani et al. (n.d.). On the opportunities and risks of foundation models (p. 3). Center for Research on Foundation Models, Stanford Institute for Human-Centered Artificial Intelligence. https://crfm.stanford.edu/assets/ report.pdf 25 Heikkilä, (2024, October 1). Why bigger is not always better in AI. MIT Technology Review. https://www. technologyreview.com/2024/10/01/1104744/why-bigger-is- not-always-better-in-ai/ 26 Pomerleau, (2022, July 26). Military services 'not aligned' on JADC2 efforts, Air Force official warns. FedScoop. https://fedscoop.com/military-services-not-aligned-on- jadc2-efforts-air-force-official-warns/ 27 John Sweller, "Cognitive Load during Problem Solving: Effects on Learning," Cognitive Science 12, no. 2 (1988): 257–285, https://doi.org/10.1207/s15516709cog1202_4 NOTES Generative AI Adoption in the US Military
Generative AI Adoption in the US Military Page 41 Page 43