OpenAI Enters Defense Pact, Filling Void After Rival's Refusal

In a significant shift, OpenAI has agreed to provide its artificial intelligence technology for use within the Pentagon's classified networks. The deal, announced late Friday by CEO Sam Altman, comes just hours after the White House ordered federal agencies to cease using services from rival firm Anthropic.

Anthropic had publicly refused a Defense Department request, citing ethical objections to potential military applications. That refusal prompted a swift response from the Trump administration, which labeled the company 'left-wing nut jobs' and mandated a transition away from its technology.

OpenAI, by contrast, stated it moved forward after the Pentagon demonstrated a 'deep respect for safety.' In a detailed statement, Altman outlined strict conditions for the partnership: the company's AI will not be used for domestic mass surveillance or to power autonomous weapon systems. The agreement explicitly requires human oversight for any use of force.

'We remain committed to serve all of humanity as best we can,' Altman wrote. 'The world is a complicated, messy, and sometimes dangerous place.'

The arrangement highlights a growing divide within the AI sector on engagement with national security. It also follows reports that Anthropic's Claude software was previously used in a military operation to abduct Venezuelan President Nicolas Maduro. Human rights groups continue to warn about the dangers of military AI, pointing to systems like Israel's 'Lavender,' used to generate targeting lists in Gaza.

For now, OpenAI has stepped into a role its competitor abandoned, betting its safeguards will hold within the world's most powerful military.

Source: Al Jazeera