The US Army is taking a bold step into the future, launching a pilot program in July that utilizes generative artificial intelligence (AI) within its acquisition, logistics, and technology (ASA(ALT)) division.

This initiative, unveiled at Defense One’s Tech Summit last Tuesday, June 18, marks a turning point in military efficiency.

It leverages the power of AI to tackle time-consuming tasks and potentially revolutionizes how the Army operates.

AI to the Rescue: Streamlining Contract Writing and Beyond

Envision a future where contract writing, infamous for its bureaucratic red tape and endless revisions, is streamlined.

The ASA(ALT) ‘s pilot program is striving for this by utilizing a Large Language Model (LLM), a unique ‘thinking machine‘ trained on the Army’s extensive data archives.

Unlike commercially available models like ChatGPT, this LLM is custom-built for military applications, ensuring the highest level of security with an Impact Level (IL) 5 classification.

This classification denotes a system authorized to store and process Controlled Unclassified Information (CUI), the backbone of many critical military operations.

“The pilot is not just about increasing our productivity,” said Jennifer Swanson, Deputy Assistant Secretary of the Army for data, engineering, and software.

She goes on to explain that the division will explore the full potential of the AI platform to test whether it will be effective across ASA(ALT) and potentially the entire Army.

This includes investigating how AI can be integrated with existing systems and potentially enhance everything from vehicles to weapon systems.

 

With its endless potential, AI algorithms have the capability to optimize logistics by predicting maintenance needs or supply chain disruptions, which could significantly enhance the Army’s operational readiness.

Building Trust and Transparency: The Human-AI Partnership

It is worth noting, however, that transparency and trust are paramount in this venture.

The LLM will provide citations for the information it generates, allowing humans to verify its accuracy and understand the reasoning behind its outputs.

A “human-in-the-loop” approach will also be implemented, ensuring human oversight throughout the process. Analysts will be able to guide the LLM, provide context for specific tasks, and ultimately make the final decisions.

This collaborative approach ensures that AI serves as a powerful tool but not a replacement for human expertise and judgment.

Mitigating Bias: Building Fair and Ethical AI

Bias, a hurdle often associated with AI, is not taken lightly by the Army.

The service acknowledges the challenge and is committed to developing ethical and responsible AI tools.

They recognize that some bias might be unavoidable, as AI algorithms are built on the data sets they are trained on.

If these data sets contain inherent biases, the AI model may reflect those biases in its outputs.

To mitigate this risk, the Army plans to employ various techniques, including using diverse training data sets and incorporating fairness checks throughout the development process.

“Bias is hard because sometimes people are biased right? So, it’s just it’s one of those very I think fuzzy things that requires oversight, and critical thinking and human involvement to make sure that we’re not just letting the tool run wild,” Swanson remarked at the Tech Summit.

The division also emphasizes the importance of critical thinking and human involvement to mitigate the impact of any potential bias.

Working on computer
Soldiers working on computer. (Image source: DVIDS)

Defend AI: Partnering for a Secure Future

This pilot program is just the first step in a larger “Defend AI” initiative.

The Army recognizes the need for collaboration with industry to develop robust and secure AI algorithms that can be seamlessly integrated into existing defense networks.

Partnering with industry leaders will ensure access to cutting-edge AI expertise and technology, accelerating the development and deployment of secure and effective military AI solutions.

While Swanson kept the LLM’s developer under wraps, she made it clear it “will not be operated by ChatGPT.

Closing Thoughts: Efficiency, Speed, and Advantage

The implications of this pilot program are far-reaching. It has the potential to significantly reduce administrative burdens, freeing up valuable manpower to focus on more strategic tasks.

Additionally, AI-powered automation could accelerate critical processes like contract negotiation and procurement, leading to faster equipment delivery and improved operational efficiency.

The Army’s embrace of AI signifies a commitment to modernization and a proactive approach to maintaining its technological edge on the global stage.

This pilot program could ignite a revolution in military efficiency, paving the way for a future powered by intelligent machines that augment human capabilities and empower the Army to defend the nation more effectively.

As AI technology continues to evolve, the possibilities for its application in the military are limitless. The US Army’s leadership in this field ensures that these powerful tools are harnessed responsibly and ethically to safeguard national security well into the future.

Disclaimer: SOFREP utilizes AI for image generation and article research. Occasionally, it’s like handing a chimpanzee the keys to your liquor cabinet. It’s not always perfect and if a mistake is made, we own up to it full stop. In a world where information comes at us in tidal waves, it is an important tool that helps us sift through the brass for live rounds.