4 weeks ago
5458
Elon Musk has spent the last 48 hours roasting and taking jabs at Amodei as the Anthropic CEO goes through what the xAI founder mocked as groveling after a public standoff with the Department of War.
Elon Musk, who has the reputation of a “world-class troll,” has not held back. He’s so far reposted clips of the interview and the leaked memo, as well as used Grok, his AI model, to generate insults and pictures of Amodei covered in mayonnaise.
Musk called the apology a pathetic attempt at groveling after getting caught. He has stated that Anthropic is “misanthropic and evil,” and accused them of training their AI to “hate Western civilization.”
Why did Dario Amodei apologize?
In a leaked internal “memo”, Amodei vented to his employees, reportedly claiming that the company’s relationship with the government soured because he refused to offer “dictator-style praise” to President Trump.
Amodei sat down with The Economist today, and he described the last few days as the most disorienting in the company’s history. He attempted to reframe the leaked 1,600-word internal document sent to over 2,000 employees as a “casual Slack post” in an effort to minimize his previous criticism of the Trump administration.
He further explained that the memo was an emotional reaction to a difficult day.
Musk’s xAI and Sam Altman’s OpenAI have already signed deals to provide their models for “all lawful purposes” without the specific constraints Anthropic demanded.
Anthropic is now offering its AI services to the Department of War at a nominal cost to prove its loyalty and utility to national security operations. Amodei stated that Anthropic has “much more in common with the Department of War than we have differences.”
Why did Anthropic get the supply chain risk tag?
For months, Anthropic has not given ground on its position not to allow its Claude models to be used by the military for mass domestic surveillance of Americans or for fully autonomous weapons systems.
The Department of War issued an ultimatum to the company ordering it to remove these “woke” guardrails, or it would risk losing its contracts.
When Anthropic refused to budge by the Friday deadline, President Trump issued a directive to “IMMEDIATELY CEASE” all federal use of Anthropic technology, and Secretary Hegseth officially tagged the company with a “supply chain risk” label.
The legal basis for the label is 10 USC 3252, a statute that allows the Secretary of War to restrict suppliers to protect the government. It is usually reserved for foreign adversaries like Huawei or Kaspersky, so its being applied to a San Francisco-based startup valued at $380 billion is a massive statement.
The label doesn’t just end Anthropic’s $200 million pilot contract; it also forces any other company doing business with the Department of War to certify that they aren’t using Claude in their military-related operations.
Anthropic has already announced it is taking the government to court, claiming the move is “legally unsound” and purely retaliatory.
If you're reading this, you’re already ahead. Stay there with our newsletter.
English (US)