OpenAI is in discussions to potentially deploy its artificial intelligence (AI) technology on NATO’s unclassified networks, as per a recent report by Reuters from a knowledgeable source. This comes shortly after the company finalized a separate agreement with the United States Department of Defense.
Initial reports by The Wall Street Journal suggested that OpenAI had been exploring the provision of services within NATO’s classified systems. However, a company spokesperson later clarified that the CEO, Sam Altman, misspoke in an internal meeting and that the potential contract pertains specifically to NATO’s “unclassified networks.”
Following the recent announcement that OpenAI would be implementing its technology within the Pentagon’s classified network, the company’s agreement with the US government came after President Donald Trump instructed federal agencies to discontinue collaboration with rival AI firm Anthropic.
Anthropic was excluded from Pentagon contracts due to reported disagreements regarding the permissible applications of its technology. The company’s CEO, Dario Amodei, has publicly expressed opposition to the utilization of its AI models for widespread domestic surveillance or fully autonomous weapons systems.
According to Reuters, the Pentagon has previously stated that it has no intention of employing AI for mass surveillance of Americans or for the development of weapons that function without human oversight but seeks to permit any lawful use of AI.
In a revised statement on Monday following the agreement reached on Friday, OpenAI declared that its AI systems would not be deliberately utilized for domestic surveillance of US citizens and nationals. Additionally, the Pentagon confirmed that the AI services would not be utilized by intelligence agencies such as the National Security Agency (NSA), as reported by Reuters.
According to The Wall Street Journal, Altman acknowledged internally that the Pentagon agreement was intricate and could pose short-term reputational risks.