Sam Altman admits rushed OpenAI's deal with Pentagon ‘looked opportunistic, sloppy’; issues clarification after backlash

Amid widescale user backlash, CEO Sam Altman has in a clarification post online acknowledged the company's rush to sign a deal with the Pentagon “just looked opportunistic and sloppy”.

A significant number of Americans cancelled their OpenAI and ChatGPT subscriptions to move to other alternatives, after Altman on Saturday announced the company had reached an agreement with the United States Department of Defense to deploy its models in classified networks.

The announcement also came amid the US government's very public feud with rival Anthropic PBC over “full military use” of its AI models and insistence on some limits. The two key sticking points for are use of AI for fully autonomous weapons and domestic mass surveillance.

Anthropic’s main app has surged to the top of Apple’s download charts in a show of support for the company during its clash with the Pentagon. The newest post from Altman is being viewed as damage control. Here's a look at the key highlights:

‘Consistent with applicable laws, critical to protect civil liberties’: Altman

In a post on social media platform X (formerly Twitter), Altman shared what he called an internal post detailing how OpenAI has been working with the Pentagon to “make some additions in our agreement to make our principles very clear”.

• At time of writing, within hours the post had amassed a million views. The OpenAI chief said the deal will be amended to add language specifying it is “Consistent with applicable laws, including the Fourth Amendment to the United States Constitution, National Security Act of 1947, FISA Act of 1978.”
• He added that it will state that AI system shall not be intentionally used for domestic surveillance of US persons and nationals.
• “For the avoidance of doubt, the Department understands this limitation to prohibit deliberate tracking, surveillance, or monitoring of US persons or nationals, including through the procurement or use of commercially acquired personal or identifiable information,” he added.

Notably, a source told Axios last month that the government refused Anthropic's demands as the categories under dispute have “considerable grey area around what would and wouldn't fall into” them, and the Pentagon is not willing to negotiate each case separately or have Anthropic's models unexpectedly block some processes.

Altman called protection of civil liberties of Americans “critical” and that the company “wanted to make this point especially clear, including around commercially acquired information”.

‘No use of AI models by NSA without follow-on contract’

• According to Altman, the Department has “affirmed” that its services “will not be used by Department of War intelligence agencies (for example, the NSA)” and that any such use “would require a follow-on modification to our contract”.

• “For extreme clarity: we want to work through democratic processes. It should be the government making the key decisions about society. We want to have a voice, and a seat at the table where we can share our expertise, and to fight for principles of liberty,” he stated.
• Altman also stated that if he “received what I believed was an unconstitutional order, of course I would rather go to jail than follow it”.

‘There are many things technology isn’t ready for'

The OpenAI chief also added that there are many things AI “just isn’t ready for, and many areas we don’t yet understand the tradeoffs required for safety”, stating that they will “work through these, slowly”, with the department, technical safeguards and other methods.

In a rare admittance he added, “One thing I think I did wrong: we shouldn't have rushed to get this out on Friday. The issues are super complex, and demand clear communication. We were genuinely trying to de-escalate things and avoid a much worse outcome, but I think it just looked opportunistic and sloppy. Good learning experience for me as we face higher-stakes decisions in the future.”

He also reiterated support for Anthropic and said it should not be designated as a supply chain risk (SCR), adding: “we hope the DoW offers them the same terms we’ve agreed to.” He first made the sentiment known on Sunday after the Pentagon's action against the CEO Dario Amodei-led company.

Notably, the ChatGPT maker and Claude AI maker's leaders have repeatedly clashed in the past over divergent approaches to AI development.

Sam Altman outlines principles: Alignment, democratization, empowerment, and individual agency

In the same thread, Altman also called the deal with Pentagon, “one of the first “real deal” decisions we have faced” and shared the principles he cared most about when making it: alignment, democratization, empowerment, and individual agency.

• According to Altman, “democratic process must stay in control” which means no private company should decide the fate of the world. “We need to work with governments, but also we need to make sure individuals get increasing power,” he added.
• “In particular, the key element required for democracy, such as protection of privacy, must be defended by all of society. I believe that, as some of the creators of this new technology, we deserve to and are obligated to have a loud voice about the risks, pitfalls, and benefits we see,” he said.
• Calling the relationship between governments and AI efforts “critical”, Altman said this will be difficult, but he does “not see any good future where we don't get there”.