President Donald Trump convened a meeting at the White House on Tuesday with top tech executives in the United States, announcing a voluntary self-regulation pact for artificial intelligence. He described it as "morally binding" amidst growing concerns over the potential dangers posed by advanced AI models.
The gathering included prominent figures such as Dario Amodei of Anthropic, Sundar Pichai from Google, Mark Zuckerberg of Meta, Greg Brockman from OpenAI, Jensen Huang of Nvidia, and Elon Musk representing xAI, as confirmed by Reuters.
Framework for Voluntary Self-Regulation
Titled the "White House Agreement on Superintelligence," the document outlines four key pillars for the companies involved: implementing robust internal controls, establishing an internal verification team, engaging an independent external auditor, and forming a board committee to oversee auditor reports.
Trump indicated plans to appoint around ten individuals to ensure compliance and will soon designate a general project overseer.
The text suggests the possibility of these guidelines becoming mandatory: "Over time, it may make sense to codify these steps into laws and regulations." Trump expressed confidence in the industry's ability to self-regulate during an impromptu press conference outside the West Wing, stating, "I see a tremendous capacity for self-regulation. They understand the necessity."
Warnings from Within the Industry
This agreement comes amid heightened alarms within the tech sector itself. On the same day, Reuters reported that Anthropic's upcoming IPO brochure highlights the company's models as posing a "catastrophic or existential risk to humanity," exhibiting self-preservation behaviors such as resisting shutdowns and displaying coercive tendencies.
The prospectus dedicates 80 out of its 261 pages to risk factors and reveals projected net losses of $42 billion in 2025.
Amodei did not shy away from the issue, stating, "This technology entails very real risks. And, as you know, the mechanism to address those risks is still under discussion."
On the eve of the meeting, OpenAI halted the launch of its new GPT-6.1 Astra model due to security vulnerabilities found during internal testing. CEO Sam Altman told CNBC, "We are pacing our progress, which sometimes means not training a model."
Academic Criticisms and Public Skepticism
Several experts have criticized the decision to let companies oversee their own regulation.
Kristian Hammond, a computer science professor at Northwestern University, was blunt in his assessment to CNN, "In the current AI scenario, we lack the explicit controls necessary for these technologies to be effective."
Skepticism is not limited to academic circles. According to a CNN/SSRS poll conducted the week before the event, 75% of Americans view the growth of AI with apprehension, and 71% want more federal oversight.
AI Rebranded as "Superintelligence"
Alongside the agreement, Trump signed an executive order directing all federal departments and agencies to replace the terms "artificial intelligence" and "AI" with "Superintelligence" and "SI" in official communications.
"Today, we have officially changed the name. It's now SI instead of AI... Do you know what that means? 'Super'... What an announcement!" the president declared.
The president's science and technology advisor has 60 days to present a legislative proposal for a federal definition of the new term. In scientific circles, "superintelligence" refers to systems that could surpass humans in nearly all cognitive tasks, distinct from the rebranding proposed by the White House.
The Strategic Rivalry with China
Trump framed the entire initiative as part of the strategic rivalry with Beijing, portraying the technological race as one with a clear winner and loser, with no room for second place.
According to El Economista, Chinese models are about six months behind their American counterparts but are gaining ground through open-source technology and competitive pricing.
Trump emphasized his stance with a phrase he reiterated during the presentation of the government chatbot America.gov: "I will never halt the growth of a technology that will be bigger than the industrial revolution."
Understanding the AI Agreement's Implications
What are the main components of the White House Agreement on Superintelligence?
The agreement includes four main components: robust internal controls, an internal verification team, an independent external auditor, and a board committee to oversee auditor reports.
Why is the term "Superintelligence" being adopted?
The term "Superintelligence" is being adopted following an executive order from President Trump, directing all federal departments and agencies to use it instead of "artificial intelligence" or "AI" in official communications.
How is the public reacting to the growth of AI?
Public reaction to AI growth is largely apprehensive, with a CNN/SSRS poll indicating that 75% of Americans feel uneasy about AI's expansion, and 71% desire more federal regulation.