The U.S. Department of Defense, also known as the Pentagon, has awarded OpenAI a one-year contract to “develop prototype frontier AI.” This contract, which will be managed through the Department of Defense’s Chief Digital and Artificial Intelligence Office (CDAO), aims to develop advanced AI prototypes to address critical national security challenges.
It is also the first awarded contract under OpenAI’s newly launched “OpenAI for Government” initiative. This initiative is designed to provide federal, state, and local agencies with secure and customized AI solutions. According to the company’s blog post, it will consolidate the tech giant’s other government-focused efforts, such as the ChatGPT Gov, which is “designed to streamline government agencies’ access to OpenAI’s frontier models,” with the U.S. Air Force Research Laboratory, NASA, Los Alamos, and the Treasury Department.
A Major Policy Shift
OpenAI’s entry into the development of military AI brings to light its previous stance and the company’s quiet policy reversal. As at the start of 2024, OpenAI’s policies explicitly banned the use of its technology for “weapons development,” as well as for “military and warfare.” However, in January 2024, the company quietly removed the explicit ban on military applications. The new guidelines now only prohibit using OpenAI’s services to “harm yourself or others,” which means that this policy change opened the door for OpenAI to pursue and/or win government and military contracts.
There’s also the complementary argument for the company’s financial and market impact. Due to the rapid and impressive growth of OpenAI, its operations are costly. The company spent about $5 billion in 2024 alone, reflecting the high expenses involved in developing and running large language models (LLMs). So it is no surprise that the tech giant changed its policies to adapt to a new system that allows the Pentagon to bring in significant revenue and also provide stable, high-margin customers in a sector that is known for its reliable budgets.
What Will OpenAI Build for the Pentagon?
According to OpenAI, the contract will support a range of applications, including:
- Improving healthcare access for service members and their families
- Streamlining analysis of program and acquisition data
- Enhancing proactive cyber defense
For the Department of Defense, they have identified AI-related projects across the military that need the expertise of a company like OpenAI. Potential uses for the company’s lucrative technology include:
- Command and control systems
- Decision support for military operations
- Operational planning and logistics
- Intelligence and cyber operations
- Autonomous and uncrewed systems
- Information operations
According to Breaking Defense, a Chief Digital and Artificial Intelligence Officer (CDAO) official noted that the contract will go beyond current large language model capabilities. Instead, it will focus on agentic workflows and semi-autonomous AI agents that can perform tasks previously handled by humans.
The Worrying and Competitive Landscape
OpenAI’s move into military AI is part of a broader trend, where other tech companies are also signing agreements and deepening their ties with governmental bodies and defense sectors.
Some of these partnerships include:
- Anthropic’s partnership with Palantir and Amazon Web Services to provide its Claude AI models to U.S. defense and intelligence agencies
- Meta making its Llama models available to defense contractors
- Authorization of Microsoft’s Azure OpenAI services to handle classified government data.
Critics of the fast-paced movement and integration of AI into our respective lives note that the lines may have been blurred as private companies are now given the mantle to make important decisions on behalf of the government. We’re steadily moving into a world many of them call “dystopian.”
Ethical Concerns and Criticism Amid a Global AI Arms Race
Some AI ethics experts and researchers have raised legitimate concerns about the potential risks of using AI in military settings. Critics argue that the integration of AI into the military could contribute to biased or imprecise military operations, which might lead to a risk increase of civilian harm. For many, this partnership between a tech giant like OpenAI and the U.S. military feels like a scene from a sci-fi thriller, where AI’s gradual rise edges ever closer to taking over the world.
Heidy Khlaaf, Engineering Director at Trail of Bits, a cybersecurity firm, pointed out in her Op-ed that shifting from an explicit ban on military applications to a compliance-based approach could, in her words, threaten, rather than preserve, US national security.
Also, ChatGPT winning the $200m Pentagon contract occurs within the context of an accelerating global AI arms race. An example is China prioritizing “intelligentized warfare,” which involves integrating artificial intelligence (AI) and other advanced technologies into its military strategy and operations.
The militarization of AI will, in no time, have profound implications for global security, especially with the absence of international governance frameworks for military AI applications, discussed and signed by all countries at the forefront of AI development.
As OpenAI continues to dominate the AI market while expanding into military applications, the contract raises fundamental questions about the marrying of AI tech companies into military bodies as well as the balance between technological innovation and ethical considerations in AI development.
