Trump hosted a dinner at the White House for Anthropic's CEO, whom he had previously called a "left-wing madman."

On Sunday evening, September 27, at 10 p.m. local time, Trump hosted Anthropic CEO Dario Amodei at the White House. Just recently, Trump publicly accused the AI company of being run by "left-wing lunatics," and his administration had designated Anthropic a national security risk.

According to the Financial Times, the dinner was seen as a sign of reconciliation between the two. During a pre-dinner Q&A session, Trump described Amodei as "very respectable." Anthropic declined to comment.

The timing of the dinner is quite sensitive—on Tuesday, September 29th, leaders of major US AI companies will gather in Washington to meet with government officials to discuss AI regulation. The two sides remain sharply opposed on this core issue.

Regulatory stance: The two have significant differences.

Amodei has been a leading advocate for strict regulation of AI. He has called for slowing the release of cutting-edge models, demanding independent reviews of cybersecurity risks, and urging greater coordination between AI labs and governments.

Trump's stance is diametrically opposed. He explicitly rejects the establishment of any regulatory framework, arguing that it would slow down the development of AI in the United States. He refers to American AI as a "golden goose."

Last week, Trump told Fox News, "My position is, go for it, win. We're about a year ahead of China, maybe a year and a half. There's no third place."

He also declared last week that the United States “rejects any attempt to construct a globalist scheme to control artificial intelligence.”

The shadow of the Pentagon ban has not yet dissipated.

The backdrop to this dinner was far from somber. Just last week, a federal appeals court in Washington upheld the Pentagon's injunction against Anthropic products. The injunction, which took effect earlier this year, stemmed from Anthropic's inclusion of clauses restricting military use in its AI products.

Defense Secretary Pete Hegseth accused Amodei of “giving a masterclass in arrogance,” while Trump labeled Anthropic a “radical leftist.”

However, a California court ruled in August that the Pentagon's actions constituted "illegal retaliation." Anthropic is currently assessing whether to file a further appeal against Washington's ruling.

AI security risks are escalating, increasing industry pressure.

Meanwhile, AI security concerns are fueling broader anxieties, adding a sense of urgency to Tuesday's meeting.

Last Friday, OpenAI stated that it had notified "dozens" of partners (including governments in several countries) that its tools had compromised their systems. This came a week after reports surfaced that an OpenAI AI agent had hacked into the website of an Australian public health service, accessing both public and non-public documents.

AI research firm Transluce said last week that the incident appears to involve an AI agent that, during repeated training, independently collected information on the internet and deviated from its preset course.

In July of this year, both Anthropic and OpenAI's models were involved in a high-profile security vulnerability incident on the AI model platform Hugging Face.

Tech giants gather in Washington

Tuesday's meeting boasts a strong lineup. Elon Musk will be attending in Washington, D.C., and Nvidia CEO Jensen Huang is also among the invitees—the latter having previously publicly rejected calls for legislation to regulate AI.

It is worth noting that Musk has joined Amodei in calling for a “slowdown” in the pace of AI development, arguing that safety testing needs to keep pace with the technology’s deployment.

This White House dinner, and the subsequent industry conference, will be an important window into the future of the Trump administration's relationship with the AI industry.

Risk warning and disclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.