According to the Financial Times, Anthropic has installed about a half-dozen engineers at the NSA, although the newspaper reports it is unclear whether Anthropic employees are actively assisting in operations against adversary countries. According to the FT, at least, they are customizing AI models for specific applications.
In March, the Pentagon labeled Anthropic a supply chain risk because the company refused to modify guardrails on the cloud. The Defense Department wanted to use the cloud without restrictions, but Anthropic wouldn’t back down on two issues: the use of its AI for autonomous weapons and mass surveillance of Americans.
Anthropic CEO Dario Amodei met with Hegseth on February 24, but they could not reach an agreement and the Defense Secretary began a crusade against him to put the company in its place. Anthropic wouldn’t back down, so President Donald Trump gave his blessing to blacklist the company, beginning the process of removing the cloud from government systems. But Hegseth’s war with Under Secretary of Defense for Research and Engineering Emil Michael doesn’t seem to be going so well.
Hegseth and his subordinates trying to argue that Anthropic was a unique threat to national security never made any sense. The company has been keen to work with the US military and its allies, as has been demonstrated repeatedly.
Amodei visited the White House in mid-April and a new report from Reuters on Friday said the dispute between Anthropic and the Trump administration is “showing signs of subsiding,” especially as the company prepares to go public. This is important for Anthropic because it makes an initial public offering safer for investors if the AI company has a good relationship with the government. But it also matters to the military and intelligence communities, as they want access to the most advanced AI in the world.
Amodei was originally invited to the White House on May 21 for Trump’s plans to sign an executive order on AI, but that was derailed after the White House decided it didn’t like some of the provisions. But it will be interesting to see the race for power among the AI giants as they side with Trump. OpenAI’s Sam Altman and XAI’s Elon Musk are having their own feud, and it can be hard to be the oligarch’s favorite when you have so many sociopathic billionaires vying for your government money.
NOTUS published a report Thursday that Altman has been considering the government taking a stake in OpenAI with Trump since the start of the president’s second term. According to NOTUS, the stake could be used “as if distributing dividend payments to all American households”, an idea that has become popular among AI enthusiasts who believe the technology will displace millions of workers and cause mass unemployment.
According to the Wall Street Journal, the US government has already taken stakes in at least ten companies, a highly unusual move usually reserved for extraordinary circumstances. The US has a 10% stake in Intel and a gold stake in US Steel. But if the government does take a stake in OpenAI it seems like it’s being pitched as something that could be used as a universal basic income or some other nominal form of financial benefit for Americans. The American people are not getting paid directly from Intel on that investment.
How wise would it be to tie the government’s fortunes to big AI companies? This remains to be seen, particularly as it potentially sets up incentives for private companies to benefit over others and ties the government’s economic stability to a specific type of technology. It may seem smart for the government to invest in the latest and greatest, but technology is constantly changing.
Imagine, for example, if the government had invested heavily in VR headsets in the 2010s and bought Oculus. The metaverse was projected as the inevitable future, so much so that Facebook changed the name of its parent company to Meta. But this gamble has not paid off in many ways, and it is a good thing that the American taxpayer is not at risk in what was always a big gamble.
Anthropic did not immediately respond to emailed questions Friday. Gizmodo will update this article if we hear back.
<a href