As debates over artificial intelligence continues to heat up, the road to AI domination is no longer simply run from Silicon Valley. It now travels through secure military servers,United Nationschamber and data centres in northern Europe.
At a recent event at theWhite Housein Washington, the US Department of Defence reportedly pressed leading artificial intelligence companies to make their most advanced systems available on classified government networks. Officials argued that faster and more capable tools could prove decisive in future conflicts.
According to sources familiar with the discussions, Pentagon Chief Technology Officer Emil Michael asked top generative AI firms, including OpenAI and Anthropic, to ensure their strongest models can operate in both classified and unclassified environments. Defence leaders believe rival powers are moving quickly to integrate AI into their military systems and that limiting access to top-tier tools could leave the United States exposed.
The push reflects a growing sense of urgency inside the Pentagon. Military planners increasingly view AI as essential for analysing intelligence, tracking threats and supporting real-time decision making. In their view, delays in adopting advanced systems could have serious consequences for both troops and civilians.
Pentagon officials have framed the effort as a matter of national defence. Reports indicate the Department of Defence has been urging companies to expand access to their systems within secure government networks so they can be used for classified missions.
'[The Pentagon] is moving to deploy frontier AI capabilities across all classification levels', an anonymous official toldReuters.
US officials are concerned that rival nations are investing heavily in AI for surveillance, cyber operations and weapons development. In that context, they fear keeping the most advanced systems confined to commercial settings could slow military readiness.
At the same time, the move raises questions about oversight. Technology companies have built safeguards into their systems to prevent misuse. Expanding their use within classified networks could reduce external scrutiny, even if internal controls remain in place.
At present, Anthropic is the only AI company whose systems are available in classified environments through third parties. Even so, the US government remains bound by the company's usage policies and does not have unrestricted access.
For many people, the debate may seem remote. Yet decisions taken now could shape how AI is used in warfare, intelligence gathering and crisis response. Some AI experts have warned about the risks of errors, including systems generating false but convincing information. In high-stakes military settings, such mistakes could have deadly consequences.
Source: International Business Times UK
