In a shocking turn of events, OpenAI’s rushed Pentagon deal has sparked a firestorm of controversy, leaving many to question the ethics of AI in warfare. But here's where it gets controversial: after CEO Sam Altman admitted the agreement looked ‘sloppy and opportunistic,’ the company is now backpedaling, promising to explicitly ban its AI from domestic surveillance and use by agencies like the NSA. And this is the part most people miss: this U-turn comes just days after Anthropic, the Pentagon’s previous AI contractor, was dropped for refusing to compromise its principles on mass surveillance, earning them a ‘leftwing nut jobs’ label from former President Trump.
OpenAI’s deal, struck almost immediately after Anthropic’s ousting, raised red flags for its 900 million ChatGPT users. Critics drew parallels to the 2013 Snowden scandal, where the NSA’s mass data harvesting exposed the dangers of unchecked surveillance. The backlash was swift, with social media users rallying behind a ‘delete ChatGPT’ campaign, one post bluntly stating, ‘You’re now training a war machine. Let’s see proof of cancellation.’ Meanwhile, Anthropic’s chatbot, Claude, surged past ChatGPT in app rankings, signaling a shift in user trust.
Altman, in a candid message to employees, acknowledged the haste, writing, ‘We shouldn’t have rushed… it just looked opportunistic and sloppy.’ OpenAI had initially claimed the deal had ‘more guardrails than any previous agreement,’ but nearly 900 employees from OpenAI and Google aren’t convinced. They’ve signed an open letter urging their leaders to resist the Department of War’s demands for AI use in surveillance and autonomous killing, warning of a divide-and-conquer tactic by the government.
But here’s the real question: Can OpenAI truly uphold its ethical promises while working with the Pentagon? Former OpenAI policy head Miles Brundage isn’t so sure. He suggests the company may have ‘caved’ to pressure, framing it as a win while undermining Anthropic’s stance. Brundage’s stance is clear: he’d ‘rather go to jail’ than comply with unconstitutional orders, advocating for democratic processes over government overreach.
As three more federal agencies drop Anthropic’s AI following Trump’s directive, the debate rages on. Is OpenAI’s amended deal a genuine step toward ethical AI, or a thinly veiled compromise? And what does this mean for the future of AI in warfare and surveillance? Let us know your thoughts in the comments—this is a conversation we can’t afford to ignore.
