We're ahead of China, and I don't want to interfere with that lead
En un momento en que la inteligencia artificial redefine el equilibrio del poder global, Donald Trump optó por la pausa antes que la firma, rechazando un decreto ejecutivo que, a su juicio, podría frenar la ventaja tecnológica de Estados Unidos frente a China. La decisión, tomada horas antes de una ceremonia en el Despacho Oval, revela una tensión más profunda: la que existe entre quienes ven en la regulación una salvaguarda necesaria y quienes la perciben como un lastre en una carrera que consideran decisiva para el futuro de la nación. En el trasfondo, modelos de IA sin publicar y sectores enteros de la economía esperan saber si Washington elegirá la velocidad o la cautela.
- Trump canceló la firma de un decreto sobre inteligencia artificial horas antes de la ceremonia prevista, argumentando que el texto podría perjudicar el liderazgo tecnológico estadounidense.
- La tensión entre la industria financiera y de seguridad —que exige regulación ante los riesgos cibernéticos— y la Casa Blanca —que prioriza la competitividad— se ha vuelto imposible de ignorar.
- El modelo Mythos de Anthropic, aún no presentado al público, se ha convertido en el centro de debates internos dentro de la administración sobre qué capacidades deben regularse y cuáles no.
- La demora no resuelve el conflicto: Trump podría revisar el decreto, archivarlo indefinidamente o usarlo como palanca de negociación con la industria tecnológica.
- Las empresas que desarrollan los sistemas de IA más poderosos del mundo observan en silencio, conscientes de que la decisión final moldeará las reglas del juego durante años.
El jueves por la mañana, Donald Trump sorprendió a Washington al cancelar la firma de un decreto ejecutivo sobre inteligencia artificial que estaba programado para ese mismo día en el Despacho Oval. La razón que ofreció fue directa: había leído el texto y no le había gustado. Su preocupación central no era técnica ni filosófica, sino estratégica. "Estamos por delante de China, por delante de todos", declaró, "y no quiero hacer nada que interfiera con esa ventaja."
La cancelación llegó en un momento de creciente presión regulatoria en Washington. Bancos, empresas de seguridad y otras instituciones han alzado la voz sobre los riesgos que representan los sistemas de IA más avanzados, especialmente en el ámbito de la ciberseguridad, donde un modelo suficientemente poderoso podría convertirse en una herramienta de ataque digital a gran escala. Pero la decisión de Trump de retirar el decreto revela dónde están sus prioridades: no en los frenos, sino en la velocidad.
En los pasillos de la administración, el modelo Mythos de Anthropic —aún sin presentación pública— ha sido objeto de discusión repetida. Sus capacidades, sus posibles usos y las reglas que deberían gobernarlos han animado debates internos que, al parecer, influyeron en la redacción del decreto ahora suspendido.
Lo que sigue es incierto. Trump podría revisar el texto y firmarlo, archivarlo sin fecha o usarlo como instrumento de negociación con la industria. Por ahora, la ceremonia está cancelada y el decreto aguarda en algún rincón de la burocracia su destino final. Las compañías que construyen los sistemas más poderosos del mundo observan, y esperan.
Donald Trump walked back a signature moment Thursday morning. Hours before he was scheduled to sit at the Resolute Desk and sign an executive order on artificial intelligence, the president announced he would not. The reason, he told reporters in the Oval Office, was simple: he had read the text and did not like what he found there.
Trump's concern was not about the order's substance in the abstract. It was about advantage. "We're ahead of China, we're ahead of everybody," he said, "and I don't want to do anything that's going to interfere with that lead." The order, as drafted, struck him as a potential liability—a constraint on American AI development at precisely the moment when the United States holds what it perceives as a decisive edge in the global technology race.
The cancellation arrived at a moment of genuine tension in Washington over how to govern artificial intelligence. Banks, security firms, and other institutions have grown increasingly vocal about the risks posed by the most powerful AI systems. These companies worry about vulnerabilities—the way advanced models can probe software for weaknesses, find them, and exploit them. The pressure for regulation has mounted accordingly. But Trump's decision to shelve the order suggests where his priorities lie: not in guardrails, but in velocity.
Behind the scenes, government officials have been closely watching Anthropic's new Mythos model. The company has not released detailed information about it to the public, but inside the administration, its capabilities have become a subject of repeated discussion. What Mythos can do, what it might be capable of doing, and what rules should govern it—these questions have animated internal debates. The executive order, apparently, would have imposed constraints that Trump now believes could slow American progress.
The delay signals a fundamental conflict that is unlikely to resolve quickly. On one side are those who see AI regulation as essential to preventing catastrophic misuse—particularly in cybersecurity, where a sufficiently powerful model could become a tool for large-scale digital attack. On the other side is the president, who sees regulation primarily as a potential handicap in a competition with China that he views as existential. Both camps believe they are protecting the country. Both cannot be right.
What happens next remains unclear. Trump could revise the order to his satisfaction and sign it. He could shelve it indefinitely. He could use the threat of regulation as a negotiating tool with the AI industry, extracting commitments in exchange for a lighter regulatory touch. For now, the Oval Office ceremony is off the calendar, and the text of the order sits somewhere in the bureaucracy, awaiting revision or burial. The companies developing the most powerful AI systems in the world are watching to see which it will be.
Notable Quotes
We're ahead of China, we're ahead of everybody, and I don't want to do anything that's going to interfere with that lead.— Donald Trump
The Hearth Conversation Another angle on the story
Why did Trump think the order itself was the problem? Couldn't he have just asked for changes?
He did ask for changes, in a sense—by rejecting it outright. But the deeper issue is that any regulation, even light regulation, creates friction. In a race, friction costs time. He sees China as the competitor and regulation as the handicap.
But the banking sector and security people are genuinely worried. Aren't they right to be?
They're worried about real things—about what these models can do to financial systems, to infrastructure. But Trump's view is that worry is a luxury America can't afford right now. If we slow down to be safe, and China doesn't, we lose.
What's the Mythos model? Why does it matter that government officials keep talking about it?
Anthropic built it but hasn't told the public much about what it can do. That silence itself is telling. Inside government, people are discussing its capabilities—which suggests it's powerful enough to matter, powerful enough to worry about. And that's exactly what regulation would try to address.
So Trump is betting that staying fast is safer than building guardrails?
Not exactly betting. He's making a choice about which danger matters more. The danger of falling behind China, or the danger of what an unregulated AI system might do. He's decided the first danger is bigger.
What happens to the order now?
It's in limbo. It could be revised and signed. It could disappear. It could become leverage in negotiations with AI companies. For now, it's a signal: Trump wants speed, not caution. Everyone else has to figure out what that means for them.