Pentagon Signs Major AI Deals With 7 Companies Including OpenAI and Google; Anthropic Remains Blacklisted Over Supply Chain and Security Concerns

The US Department of Defence has formalised agreements with seven major artificial intelligence companies to integrate their technology into classified military networks. This expansion aims to diversify the Pentagon’s AI providers and prevent over-reliance on a single developer. While industry leaders such as OpenAI, Google, and Microsoft have secured spots on the military’s secret and top secret networks, Anthropic remains conspicuously absent. The startup continues to be barred from the department due to an ongoing dispute over safety guardrails and supply chain concerns.

Pentagon New Partnerships for Classified Networks

The Pentagon confirmed on Friday that SpaceX, OpenAI, Google, NVIDIA, Reflection, Microsoft, and Amazon Web Services (AWS) will now provide AI capabilities across sensitive defence environments. These tools are used by military personnel for a range of tasks, including strategic planning, logistics, and targeting. By incorporating multiple vendors, the Pentagon stated it hopes to avoid "vendor lock," a situation where the military becomes too dependent on one company’s proprietary technology. Anthropic Explores In-House AI Chip Design Amid Rising Demand and Hardware Shortages: Report.

Notably, the list includes Reflection AI, a newer entrant that recently raised $2 billion. The startup has garnered attention due to its backing by 1789 Capital, a venture firm where Donald Trump Jr. is a partner. The inclusion of smaller or newer firms suggests a significant shift in how the US military procures advanced technology.

The Ongoing Conflict With Anthropic

Despite being widely used by individual personnel within the Department of Defence, Anthropic was officially labelled a supply chain risk in March. This designation has led to a legal battle between the company and the Pentagon. The disagreement stems from the guardrails Anthropic places on how its AI models can be used in military contexts.

Pentagon Chief Technology Officer Emil Michael reiterated on Friday that Anthropic remains a risk. He pointed to "Mythos," the company’s latest AI model, as a specific national security concern due to its advanced cyber capabilities. While many organisations are using previews of Mythos to defend their IT infrastructure, US officials remain wary of its potential to "supercharge" hacking efforts.

Accelerated Integration and Personnel Demand

Since the blacklisting of Anthropic, the Pentagon has significantly reduced the time it takes to vet and deploy new AI software. The process of moving a company onto top secret data levels, which previously took up to 18 months, has been shortened to less than three months. This speed is essential for the military’s main AI platform, GenAI.mil, which has reportedly been used by over 1.3 million personnel in just five months of operation.

However, the transition away from Anthropic’s tools has met with internal resistance. Several Pentagon staffers and IT contractors have expressed a reluctance to stop using the startup’s software, which they consider superior to current alternatives. Despite these views, the military has issued orders to remove all Anthropic tools within the next six months.

Anthropic Situation May Not Be Permanent

The current exclusion of Anthropic may not be permanent. Last week, President Donald Trump suggested that the company was "shaping up" in the eyes of his administration. This comment has opened the possibility that the blacklisting could be reversed if the company reaches a compromise with the government regarding its security protocols and usage restrictions. Google-Anthropic Deal: US Tech Giant Eyes USD 40 Billion AI Investment As Competition in Advanced Models Intensifies.

For now, the Pentagon is moving forward with its seven chosen partners. Google has already signed a specific deal to allow its models to be used for classified work, and the integration of NVIDIA and SpaceX ensures that both hardware and satellite connectivity are deeply embedded in the US military's AI strategy.

(The above story first appeared on LatestLY on May 03, 2026 01:09 PM IST. For more news and updates on politics, world, sports, entertainment and lifestyle, log on to our website latestly.com).