[Phone] The President of the United States, Trump, had a private dinner with the chief executive officer of the artificial intelligence company Anthropic at the White House on Sunday night. This was the first time that they had met separately after months of intense confrontation, in the context of the Federal Court of Appeal's ruling two days ago that the Pentagon had legalized the designation of Anthropic as a "risk to the national security supply chain".
The dinner was personally invited by Trump. According to several media reports, Amody was unable to attend a national dinner for President Xi Jinping at the White House last week due to scheduling conflicts. – The national dinner invited such scientific and technological figures as OpenAI CEO Sam Otmann, Amazon Founder Bezos, and CEO Zuckerberg of the United States dollar, but Amody was not listed. Trump told journalists at the Andrews joint base in Maryland that he would have dinner with Amody at 10 p.m. that night and described Amody as "respectable".
It is worth noting that just before this meeting, a black memo against Amody was circulated within the White House. The memorandum lists the positions held by the Amody and Anthropic teams — including abortion, veganism, over-emergence of animal values, civil rights protection for the “digital consciousness” class — but does not directly link these concepts to Amody himself. According to sources, White House officials had advised Amody not to seek a meeting with the President, saying, "Dario is a little too strange for Trump."
The beginning of this confrontation dates back to 27 February. On the same day, Trump instructed the federal agencies to "stop immediately all technologies using Anthropic" and set a six-month transition period for institutions such as the Ministry of Defence. The Minister of Defence, Peter Hegessys, immediately announced the designation of Anthropic as a "national security supply chain risk". On 5 March, Amody acknowledged receipt of the designated letter from the Ministry of Defence, which became effective immediately. Since then, federal agencies have taken action to strip the Claude model of Anthropic.
At the heart of the conflict is the refusal of Anthropic to remove security barriers associated with large-scale surveillance and autonomous weapons from its products. Amordi is clearly opposed to the use of Anthropic technology for these purposes, while Trump advocates accelerated development of AI, arguing that the United States must retain its advantage in the science and technology competition with China. According to Trump, the United States is leading China in the AI competition for about a year and a half.
Following a further escalation of the dispute, the Minister of Commerce, Howard Lutnik, threatened to bring criminal charges against Anthropic, prompting criticism by scientific and technological groups and lobbying by the British and other allies. The Department of Commerce also ordered Anthropic to obtain United States approval before foreign nationals could use its Fable 5 and Mythos 5 models.
Just two days before the dinner, the Federal Court of Appeal ruled that the Pentagon had legalized the listing of Anthropic as a supply chain risk and allowed the Ministry of Defence to continue removing the Claude model of Anthropic from its system. This decision should have further consolidated the Administration ' s position, but Trump chose to sit down with Amody.
In fact, Trump's attitude towards Anthropic has been significantly softened in recent weeks. In previous pre-recorded interviews with various media reports, Trump stated that Anthropic "is no longer a threat to national security" and praised the company for "showing very responsible behaviour". When asked if he thought Anthropic was a threat, he said, "Well, not now. But maybe a week ago. "He also expressed his willingness to relax, but "I'm not sure it's necessary."
However, formal federal restrictions remain in force.
The meeting took place against the background of the gap between the pace of development of AI and security research. Amody wrote earlier this month to appeal to AI to consciously slow down the development of increasingly powerful models on the grounds that safety research has not kept pace with technological progress. He wrote, "We have to slow down the upgrade of the AI model. Progress still seems to be fast, and we must use the time we have won wisely." Previously, a researcher from Anthropic resigned because of the serious risks that AI posed to humanity.
By contrast, on September 14, in the "All In" podcast, Trump mocked Amordi as a "perfect little angel" and criticized his position calling for the AI guard. The memorandum circulated within the Trump administration also reflects deeper philosophical and political differences — the White House and its allies view Anthropic as an institution full of Biden liberals, Democrats and “unusual Californians”.
At the end of the dinner, Trump is expected to issue a statement from the elliptical office at 2 p.m. Eastern Monday time. The central concern is whether the meeting will lead to a formal easing of federal restrictions on Anthropic.
During the G7 Summit (held in Evian Leban, France), Amody met again with Trump and Anthropic technicians held separate consultations with government officials. Trump stated at the time that Anthropic had "responsiblely shown" but that formal restrictions remained.
For the AI industry, the White House dinner sent a clear signal that, despite the court ' s support for the Ministry of Defence ' s decision to seal off, and despite the fundamental differences between the two sides over the AI security concept, the Trump government was willing to engage in direct dialogue with the head AI company. For Anthropic, this was a breathing opportunity after months of legal warfare and administrative pressure, but it remains to be seen whether it can be translated into real policy easing.