An inside look at DOD's AI strategy

The Defense Department is on the precipice of implementing AI and machine learning technologies, which could usher in a game changer for delivering capabilities to your biggest customer.

The Department of Defense stands on the precipice of implementing artificial intelligence (AI) and machine learning technologies. AI promises to be a game changer in delivering capabilities that will increase combat effectiveness—everything from tactical warfighting to ensuring readiness through enhanced logistical operations. The next few years will be a critical period for determining where and how AI can be adopted by the DOD.

While there are few places within the DOD currently operationalizing AI, there are many that are evaluating it. There is no shortage of use cases, and once adoption takes off, growth will be exponential.

Project Maven, the DOD’s AI solution for analyzing imagery for intel purposes, has to date been the biggest AI project, and funding has increased from $16 million in fiscal 2018 to $93 million in fiscal 2019 — a 480 percent increase. As the DOD pilots AI projects for things like improving maintenance and repair of weapons systems, supply chain management and improving business processes, industry can expect to see similar rates of growth.

The central organization within DOD that will be critical to AI implementation is the Joint Artificial Intelligence Center (JAIC). JAIC was created quickly to ensure that DOD effectively and ethically builds out its AI capabilities. The organization will look at AI cross-domain solutions across the Service branches, as well as specific component projects. Use cases that JAIC is looking at include solutions for humanitarian assistance, ISR, GEOINT, force protection, cyber and robotic process automation. (One use case that will be low hanging fruit for DOD is using AI to improve logistics and supply chain management.)

With AI poised to change DOD-wide operations, technology companies whose solutions are not specific to AI or machine learning may be left wondering how they can participate in such an exciting (and growing) technology trend.

At its core, AI is really about data—data quality, collection, integration, cleansing, protection, storage—not to mention data architecture that can support and enable AI. JEDI (Joint Enterprise Defense Infrastructure), DOD’s planned enterprise cloud, is often mentioned alongside AI, as DOD will need to depend on having the right infrastructure set up to make the data needed for AI algorithms accessible.

Technology companies should frame their solutions as AI enablement tools, whether that’s foundational in terms of laying the infrastructure for an architecture that better supports AI; assisting with data integration, quality and collection (often the most challenging piece of any analytics effort); or improving data accuracy and reporting.

Technology companies should also note that DOD will rely on commercially available technologies as much as possible when it comes to AI, as they are still working on growing in-house expertise. JAIC has a start-up mentality, looking to OTAs (Other Transactional Agreements) as ways to quickly scale and prototype AI projects. According to DOD CIO Dana Deasy, JAIC is “open for business,” so the time is ripe to get involved with their projects.

Two other DOD organizations that will be important for AI are DARPA (Defense Advanced Research Projects Agency) and Defense Innovation Unit, which will both be working AI projects alongside JAIC. Service branches will also be looking at piloting AI efforts of their own in the next year, especially in supply chain and logistics management.

There are still many considerations when it comes to AI implementation. Along with the more obvious ethical implications AI raises in warfighting, DOD leaders also need to consider how to buy AI at speed and scale. All technology companies, regardless of what type of technology they focus on, will need to pay attention to the development of AI. It will affect traditional infrastructure and cloud, edge computing and IoT, cyber data collection and protection, and of course, analytics and business intelligence.

AI has the potential to completely modernize how DOD’s data is used across every mission. Whatever type of technology or services you offer, consider how it can help DOD achieve AI success by getting your customers ready for the coming data revolution.