Pentagon signs AI deals with eight tech giants for use on military’s classified networks"
In its biggest-ever effort to integrate private AI into its operations, the U.S. Department of Defense said yesterday it had agreed to test and implement the latest and most advanced artificial intelligence (AI) technologies by eight leading vendors on classified military networks.
This is part of the Department's efforts to create what it calls an "AI-first" military — using AI to drive decision-making across all areas of warfighting — without giving control over the development or operation of those systems to any individual vendor.
As reported by CNN Business last year, the Department signed an agreement with Google Cloud and Microsoft Azure to explore integrating commercial AI technologies into classified systems. However, according to a report by Bloomberg News published today, the agreements do not include either Google Cloud or Microsoft Azure, nor does the report indicate that any other specific company has been named as having signed the agreement.
According to the report, the Department has not disclosed which vendors were selected to participate in the project, nor has it provided any details regarding how they plan to ensure that the AI systems will operate safely and securely.
The Department stated that it plans to use these new technologies for mission-critical applications and to enhance its ability to gather intelligence and make decisions quickly. However, many experts believe that deploying untested AI systems in critical situations could result in unintended consequences.
Get the latest model rankings, product launches, and evaluation insights delivered to your inbox.
While the exact timeline for the implementation of this program has not been disclosed, according to the report, the Department expects that the first implementations will begin later this year.
According to a report by Politico, this move represents a significant shift toward embracing private sector innovation in national security, however many lawmakers remain skeptical. According to Politico, some lawmakers question whether the Department of Defense is sufficiently prepared to manage and regulate the proliferation of AI systems across multiple sectors of the DoD.
According to a report by Reuters, the Department of Defense stated that it will also take steps to limit the potential for "vendor lock," where a particular vendor becomes too dominant in terms of AI system design, procurement and maintenance. According to Reuters, "Vendor Lock" has long been a concern in the government contracting space, particularly after Microsoft lost out on a $10Bn cloud computing contract in 2019 due to accusations of favoritism towards Amazon Web Services.
According to a report by C4ISRNET, some experts say that the Department of Defense should be taking a more cautious approach.
"The DOD needs to slow down and think through the implications of this rapidly evolving field," said Dr. Robert Work, former Deputy Secretary of Defense.
"AI is changing so fast that we don't even know yet what kind of threats it poses to our country."
Work told C4ISRNET that he believes that there are many promising uses for AI within the Department of Defense, including analyzing large amounts of intelligence data and making quick decisions during times of crisis.
However, he also expressed his belief that the Department of Defense should proceed cautiously because it lacks a clear understanding of how these systems might interact with each other and how they might affect future missions.