and contracting, AI can analyze supplier performance, draft procurement docu- ments, and assess risks. In testing and evaluation, AI can simulate operational environments, optimize test plans, and analyze large volumes of test data. Life- cycle management benefits from predic- tive maintenance models, sustainment planning, and digital twins for real-time monitoring. AI can also generate realis- tic training scenarios, synthesize docu- ments, and analyze historical data for knowledge management. In addition to applications high- lighted here, generative AI can pro- pose alternative solutions and even create initial drafts of the acquisition strategy. AI can “score” a draft strat- egy by grading the document, sug- gesting improvements, and identify- ing possible gaps. If used effectively, AI can save a lot of time in drafting the document and improving the strat- egy. However, we must distinguish between developing and writing the strategy. Developing the Acquisition Strategy Strategy is one of the IPT’s most important tasks. Regardless of which pathway (or pathways) is selected, the acquisition strategy must inte- grate all of the program functional areas, risks, opportunities, and con- straints into a comprehensive plan of action. Once the strategy is ap- proved, it is difficult and costly to
using predictive AI to predict trends in obsolescence, material availability, supply chains, and reliability, IPTs can make informed decisions on strategy affecting the maintenance concept, data rights, and even the system’s technical architecture. Predictive AI can also help to iden- tify framing assumptions that are key suppositions central to shaping cost, schedule, and performance expec- tations. These critical assumptions, usually just a few, are documented in the acquisition strategy. Root cause analyses show that erroneous fram- ing assumptions have resulted in sig- nificant acquisition baseline breaches that include cases where programs greatly exceeded their schedules and budgets. Since AI can predict trends with new technology advancements and other future outcomes, IPTs can use the tool to derive more accurate assumptions, enabling credible ex- pectations at the program outset. Generative AI A different and exciting dimen- sion opens with generative AI, in which advanced computer programs analyze and use data to generate new content. Generative AI also relies on access to huge amounts of data. Thus, the quality of the new content can be related to the amount of data that the software can pull from. This can be a limiting factor for acquisition because of restricted access to sensitive data. Here is what AI says about using generative AI in acquisition in re- sponse to the same prompt used about predictive AI: “In what ways could generative AI be used in defense acquisition?” Generative AI has the potential to revolu- tionize defense acquisition processes by enhancing efficiency, decision-making, and innovation. It can automate require- ment drafting by analyzing historical data and mission needs, identify gaps in current capabilities, and simulate operational scenarios to ensure mis- sion-critical requirements are met. In concept development and prototyping, AI can generate multiple designs, create digital prototypes for rapid testing, and optimize performance. For procurement
change since so many efforts are set in motion, including the Request for Proposal, Source Selection Plan, and the various functional plans that align with the strategy (e.g., Systems Engi- neering Plan, Product Support Plan). Thus, it is critical to develop a sound strategy at the program’s inception. Developing an effective strategy requires critical thinking in analyzing information, developing assumptions, and forming judgments. This includes evaluating different alternatives be- fore selecting an approach. That evaluation involves context, priorities, stakeholder interests, professional judgment, and many other factors that likely will not reside in any AI- produced plan. One could argue that AI makes critical thinking even more important as the access to content accelerates both in scale and in time. AI has made significant progress in a short period of time, but its ability to think critically is still under debate. While AI can perform the predictive and generative tasks, it cannot deter- mine all the surrounding background information associated with each pro- gram’s unique circumstances. This is where human judgment is needed to properly guide the thinking and ma- chine learning outputs. Innovation or creative thinking is another aspect of strategy devel- opment because we should use the
AI has made significant progress in a short period of time, but its ability to think critically is still under debate. While AI can perform the predictive and generative tasks, it cannot determine all the surrounding background information associated with each program’s unique circumstances.
30 | DEFENSE ACQUISITION | May-June 2025
Made with FlippingBook - Online Brochure Maker