News broke Monday night that OpenAI and the Pentagon have amended their controversial deal to include more words about privacy protections. According to reporting from Axios, the following lines were added:
- “Consistent with applicable laws, including the Fourth Amendment to the United States Constitution, National Security Act of 1947, FISA Act of 1978, the AI system shall not be intentionally used for domestic surveillance of U.S. persons and nationals.”
- “For the avoidance of doubt, the Department understands this limitation to prohibit deliberate tracking, surveillance, or monitoring of U.S. persons or nationals, including through the procurement or use of commercially acquired personal or identifiable information.”
The significance of these specific changes is not hard to trace. A New York Times story from yesterday purported to detail exactly what has caused the rupture between the Pentagon and OpenAI rival Anthropic—culminating in Anthropic being designated a “supply-chain risk” and barred from doing business with many major companies.
Essentially, that times reports says, Anthropic spoke up about surveillance involving certain kinds of unclassified bulk data on Americans that can track people’s physical location and browser histories. The final break in the negotiations stemmed from Anthropic’s request for what the Times called a “legally binding promise from the Pentagon not to use its technology on unclassified commercial data.”
The Pentagon has maintained throughout this process that Anthropic has asked for provisions requiring the Pentagon not to do things that are already illegal. Pentagon spokesman Sean Parnell wrote on X that “The Department of War has no interest in using AI to conduct mass surveillance of Americans (which is illegal).” The Pentagon merely wants to be granted the right to do anything legal, Parnell claims, which would “prevent Anthropic from jeopardizing critical military operations and potentially putting our warfighters at risk.”
OpenAI’s Sam Altman claims to share Anthropic’s concerns. And according to Altman’s X post from Monday night about the latest negotiations, it sounds like there’s been a lot of back and forth about this—evidently with the Pentagon continuing to stress that mass surveillance is ostensibly already illegal, and with OpenAI stressing that the Pentagon nonetheless still has to actually be constrained by the terms of the deal.
Here is re-post of an internal post:
We have been working with the DoW to make some additions in our agreement to make our principles very clear.
1. We are going to amend our deal to add this language, in addition to everything else:
“• Consistent with applicable laws,…
— Sam Altman (@sama) March 3, 2026
Although it would be speculative at this point to say there’s been any sort of material cost to OpenAI after it signed its Pentagon deal on the eve of the latest U.S. military action against Iran, it would be perfectly fair to say people have gotten pretty mad at the company.
There’s now a website called QuitGPT, calling for a boycott of ChatGPT. The homepage has a little counter claiming without any sort of citation that 1,513,922 people (as of this writing) have joined the boycott. The site says participants can “make an example of ChatGPT,” and “send a clear signal to ICE enablers that their actions will not go unpunished.” This doesn’t really correspond to any tangible difference between what Anthropic and OpenAI have been allowing the government to do with their respective products, but it certainly follows from Donald Trump dubbing the folks at Anthropic “leftwing nut jobs.”
Oh, and Katy Perry has announced that she has switched to Claude for all her AI needs. So clearly times are tough for OpenAI.
Gizmodo reached out to OpenAI for information about any effects from this apparent backlash, or any comment the company would like to provide about it. We will update if we hear back.
Read the full article here
