Trump and Xi endorse the term “superintelligence” as Meta invests in a billion-dollar lab

The new term proposed by the United States

On Friday, September 25, US President Donald Trump announced that Chinese President Xi Jinping had agreed to adopt the term “superintelligence” instead of “artificial intelligence.” In a post on Truth Social, Trump described the meeting as “very productive” and said that “extraordinary things are coming.”

Trump wrote: “The meeting was very productive with Xi, both for the US and for China. Extraordinary things are coming. Like almost everyone, he seemed to like calling artificial intelligence… a much more accurate and important name: SUPERINTELLIGENCE.”

A week earlier, during a speech at the UN General Assembly on September 22, Trump had already decided that the US government would use “superintelligence” (SI) in all official documents. The president argues that the word “artificial” distances the technology from humans, while “super” reinforces the idea of a tool at the service of people, an attempt to neutralize the negative connotation associated with the original term.

Political impact and debate about risks

The name change comes amid a public debate about the existential risks of AI. Trump dismissed warnings from industry executives such as Dario Amodei (Anthropic) and Sam Altman (OpenAI) as “radical left-wing rumors,” while the two CEOs called for stricter international controls at the UN Security Council.

The president also said that the Department of Justice would monitor illegal uses of the technology, but avoided proposing new regulations. Analysts link this stance to fears that slowing AI development would put the United States at a disadvantage in the technological competition with China. The issue dominated Xi’s state visit to Washington, which included a dinner at the White House attended by leaders such as Mark Zuckerberg, Altman, and Elon Musk.

Meta and its billion-dollar bet on superintelligence

Meta created the Meta Superintelligence Labs (MSL) on June 30, 2025, after investing $14.3 billion in a stake in Scale AI to bring its founder, Alexandr Wang, then 28, as AI director—the first position of its kind in the company’s history.

The lab recruited top researchers from competitors such as OpenAI, Google DeepMind, and Anthropic, reaching around 3,000 employees. An elite team, called TBD Lab, was formed to develop next-generation models, including Muse Spark, which powers the Muse personal assistant.

In March 2026, Meta established a parallel “applied AI engineering” unit under CTO Andrew Bosworth, which reduced Wang’s autonomy and signaled a greater priority on immediate revenue products over long-term research.

In October of the same year, the company cut about 600 jobs from the lab, affecting fundamental research, applied AI, and infrastructure areas, but preserving the TBD Lab. Wang remains at the helm of the division, and Meta continues to hire for its most advanced research core, including professionals from OpenAI and Thinking Machines, a startup founded by former OpenAI CTO Mira Murati.

Mark Zuckerberg used the term “superintelligence” again in September, explaining the long-term vision behind the launch of Muse, keeping the word in the company’s official vocabulary despite internal turmoil.