OpenAI is indeed creating a super agent for government uses, and the details have been released after a closed-door meeting.
The rumor mill went into overdrive ahead of the meeting a week ago with speculation that OpenAI was showing something special to the US government officials invited to come along.
Now, OpenAI has unveiled ChatGPT Gov and claims that it will allow Government workers to feed “non-public, sensitive information” into OpenAI’s models, but with the AI operating within secure government hosting environments.
What Is ChatGPT Gov?
This AI agent was built specifically for government use, according to CNBC, and the emphasis, unsurprisingly, is on security.
OpenAI CPO Kevin Weil briefed reporters on the new platform explaining that it looks like ChatGPT Enterprise, but government workers will use it in their own Microsoft Azure commercial cloud or Azure government community cloud. This means that they can “manage their own security, privacy and compliance requirements,” added OpenAI’s Felipe Millon.
This just in! View
the top business tech deals for 2025 👨💻
The briefing also included a demo of how a government employee will use the AI tool on a day to day basis. Tasks included creating a five-week plan of their tasks and then analyzing an uploaded photo of this plan but with notes and annotations all over it.
From this, it created a memo summarizing the tasks listed, which was to be sent to the legal and compliance department. And, as a final flourish, the memo was translated into different languages.
Due Diligence
The briefing also included an update on how ChatGPT Enterprise is going through the Federal Risk and Authorization Management Program and it can’t be used for sensitive data until it has got this accreditation.
OpenAI CPO Kevin Weil said that this could be a “long process” but added: “I know President Trump is also looking at how we can potentially streamline that, because it’s one way of getting more modern software tooling into the government and helping the government run more efficiently.”
Because of this, though, there is no timeline for the rollout of ChatGPT Gov, with Weil just saying it would arrive in the “near future.” He added that testing and live use could happen “within a month.”
Safety Concerns
Weil stated that the new agent will have the biggest impact in defense, law enforcement and health care, but these are also the areas where the data is potentially the most sensitive. However, there are concerns about AI safety and the increasing cozy relationship between OpenAI’s CEO, Sam Altman, and Donald Trump is doing nothing to allay fears.
Ahead of Trump’s inauguration, two senators – Elizabeth Warren of Massachusetts and Michael Bennet of Colorado – even published an open letter accusing the bosses of the big tech companies of trying to court favor.
The senators suggested they had made “million-dollar gifts to President-elect Donald Trump’s inaugural fund in what appears to be an effort to influence and sway the actions and policies of the incoming administration.” Altman was among those who donated and attended the ceremony.
Directly addressing the fact that all of those who donated are “currently under scrutiny from federal regulators,” the senators said: “You have a clear and direct interest in obtaining favors from the incoming administration: your company and many other Big Tech donors are already the subject of ongoing federal investigations and regulatory actions.”
Altman retorted in response on X: “Funny, they never sent me one of these for contributing to democrats.”
The AI Arms Race Heats Up
Weil absolutely set the tone when he said that ChatGPT Gov – and indeed all of OpenAI’s innovation – is underscored by “how important it is that the US wins this race.” With the unveiling of Trump’s Project Stargate, lauded as the “largest AI infrastructure project in history,” AI is an absolute priority for this administration.
One of Biden’s last acts was a bid to secure energy sources for new AI infrastructure. His approach has been slow and steady, pushing for investigation into the potential issues that AI could bring.
In contrast, on Day One, Trump overturned a previous executive order designed to mitigate the risks of AI and is taking a build-first, think-later approach to AI.
China is firmly in his crosshairs and he is determined to win this race – the Space Race of our age – but will there be any consideration of potential dangers as innovation is pursued at a breakneck speed?
