The US Department of Defense has reached an agreement with seven artificial intelligence companies to use advanced artificial intelligence systems in secret and top-secret military networks. The Pentagon’s new agreement list included SpaceX, OpenAI, Google, NVIDIA, Reflection AI, Microsoft and Amazon Web Services. Anthropic was left off the list. Pentagon is expanding its artificial intelligence supply chain. According to Pentagon’s statement, SpaceX, OpenAI, Google, NVIDIA, Reflection AI, Microsoft and Amazon Web Services will be included in the sensitive network environments of the Department of Defense.
Some of these companies were already working with the Pentagon. With the new process, military personnel will gain wider access to the products of these companies on sensitive issues. Artificial intelligence tools are used in the US army in areas such as planning, logistics, targeting and faster execution of large operations. The Pentagon’s main AI platform, GenAI.mil, has been used by more than 1.3 million Department of Defense personnel within five months of its launch.
This figure shows that artificial intelligence is no longer a limited field of experimentation in the US defense structure. The most striking aspect of the agreements was Anthropic’s exclusion. Anthropic had a dispute with the Pentagon over safety limits for military use. The Department of Defense identified Anthropic as a “supply chain risk” earlier this year and banned the use of the company’s products by the Pentagon and its contractors.
There is also resistance within the Pentagon to abandoning Anthropic tools altogether. Speaking to Reuters, Pentagon employees, former officials and IT contractors who work closely with the US military stated that they see some Anthropic tools as superior to alternatives. Despite this, the Pentagon ordered that Anthropic products be removed from systems within six months. Anthropic’s exclusion increased the Pentagon’s interest in other artificial intelligence initiatives and large technology companies.
Following the tension between the Ministry of Defense and Anthropic, the process of recruiting new artificial intelligence companies to secret and top-secret data levels has also accelerated significantly. The integration process, which previously took 18 months or more, has decreased to less than three months, according to some new companies. With new agreements, Pentagon reduces the risk of dependence on a single provider, known as “vendor lock”.
This statement directly conveys the message of not being overly dependent on Anthropic or another dominant AI vendor. The Ministry of Defense increases operational flexibility by providing military units with access to different artificial intelligence systems. Reflection AI, which was on the list, was one of the noteworthy companies. The lesser-known startup raised $2 billion in October. Among the investors of Reflection AI is 1789 Capital, where Donald Trump Jr.
is a partner and investor. There is also a separate development on the Google side. A source speaking to Reuters stated that the Pentagon signed an agreement that allows it to use Google artificial intelligence models in secret missions. Google was already among the technology providers used by the Pentagon. On the Anthropic front, risk assessment is not completely closed. Department of Defense Chief Technology Officer Emil Michael told CNBC that Anthropic is still considered a supply chain risk.
Michael stated that, on the other hand, Anthropic’s artificial intelligence model called Mythos, which has advanced cyber capabilities, creates a separate national security topic. Mythos has become a tool that attracts attention among US officials and companies due to its capacity in the field of cyber security. Numerous public and private sector organizations have received access to the preview version of Mythos to strengthen their infrastructure against future cyber attacks.
It is not clear whether the Pentagon is included in this preview program. US President Donald Trump also made a remarkable statement about Anthropic last week. Trump said the company was “beginning to take shape” under his administration. This statement was interpreted as Anthropic’s blacklist status in the Pentagon may change in the future. However, in the current table, the company is still not on the agreement list and is not among the Pentagon’s new artificial intelligence partners.
Pentagon’s latest move points to a new era in the use of military artificial intelligence. Major technology companies and next-generation artificial intelligence startups are now being directly integrated into covert military networks. Anthropic’s exclusion shows that this process is shaped not only by technological capacity, but also by military use limits, supply security and political considerations.
Comments
You can write your views about this story. Comments may be moderated according to site settings.