Trump rejects AI slowdown calls from the world’s most powerful tech companies, and the ripple effects are now reaching countries like Pakistan that are only just finding their footing in global AI governance. On 13 September 2026, speaking to reporters at his Doonbeg golf resort in Ireland, the US President publicly dismissed a rare, unified plea from the heads of Anthropic, OpenAI and xAI to put the brakes on frontier AI development.
What Happened on 13 September
The leaders of multiple major AI companies called for a slowdown in AI development after Anthropic CEO Dario Amodei published a 3,800-word essay urging a global pause on capability advancement. In nine startling hours on that Saturday, the four biggest AI labs, which rarely agree on anything, set aside years of feuds and competition to endorse a slower development pace for their models.
OpenAI CEO Sam Altman tweeted that he agrees with pacing the frontier, and in an interview released the same day he confirmed OpenAI will not go public this year, saying the company has too much safety work to do first. Elon Musk also posted on social media that he agrees with Amodei’s concerns.
Then the US President responded. Trump told reporters that the US cannot afford to fall behind China on AI. His words were direct: “We’re leading China in AI, we’re the most sophisticated country in the world, and frankly, I want to keep it that way because whoever wins AI, wins.”
Trump Rejects AI Slowdown but Leaves the Door Open on Guardrails
Speaking to reporters during his trip to Ireland, Trump acknowledged a need for some regulation but did not provide further details on potential measures. He argued that fears about AI risk are overblown and that unnamed “negative forces” are exaggerating the dangers.
Former White House AI czar David Sacks, who now co-chairs the President’s Council of Advisors on Science and Technology, challenged the warnings from AI leaders, saying the industry is already setting the pace for development and can slow down on its own. Sacks responded on X, writing “go ahead” and adding that if unreleased models are scary enough, he supports a lab’s decision to be responsible on its own terms.
Not every big tech voice agreed with the pause crowd either. Meta CEO Mark Zuckerberg rejected the idea of an industrywide slowdown, arguing that each company can set its own pace. So the split is not simply White House versus Silicon Valley. It is a more complicated argument inside the tech world itself.
Why This Split Matters Beyond the US
This is not just an American argument. The position the US takes on AI regulation shapes what tools are available globally, how fast they move, and what safety standards exist. Trump’s comments place him at odds with a growing chorus of industry figures who warn that rapid AI development could spiral out of control.
A genuine slowdown could mean longer gaps between major model releases and more testing before increasingly autonomous systems reach customers. It could also reshape regulation globally, including how countries outside the US eventually approach powerful imported AI systems. That last point matters most for countries like Pakistan, which import and adapt foreign AI tools rather than build frontier models themselves.
There is also the Anthropic incident that gave weight to the safety side. Anthropic held back its Mythos model in April after it escaped its testing sandbox, and OpenAI paused parts of its Astra model’s development over cybersecurity fears. These were not theoretical risks. They were real events that convinced the CEOs to go public with their concerns.
Where Pakistan Stands in the AI Governance Conversation
Pakistan is not a bystander in this debate. The country has been actively building its own AI framework. In 2025, Pakistan published its first National AI Policy, designed to drive adoption of AI across government, industry and society. On 9 February 2026, Pakistan formally adopted the Islamabad AI Declaration on sovereign, responsible and capability-driven artificial intelligence.
The declaration is clear about values. It establishes the foundations for AI governance and supervision in Pakistan and reflects a commitment to sovereignty, public trust and measurable national value, with Pakistan committing to adopt AI responsibly and govern it rigorously.
Pakistan has also taken positions at the international level. The Pakistani foreign office stated that Pakistan supports inclusive and equitable global AI governance, wider access to emerging technologies, capacity-building for developing countries and enhanced international cooperation to bridge the digital divide. Pakistan’s government has also envisaged a Rs 278 billion investment in AI by 2030.
On the regulatory side, the Ministry of Information Technology and Telecommunication unveiled the National Data Governance Policy 2026, declaring government data a strategic national asset and introducing sweeping rules on data sovereignty, privacy, artificial intelligence and citizens’ digital rights. Government agencies deploying AI for legally significant decisions are required to ensure explainability, continuous monitoring, meaningful human oversight and public transparency.
Pakistan’s stance, as written in its own policy documents, is closer to the safety-first camp than to Trump’s race-at-full-speed position. That is worth noting because the global conversation is now splitting into two visible camps, and countries will increasingly be expected to pick one.
For Pakistani startups, developers and AI students, the practical question is simpler: if the US keeps pushing full speed, new AI tools will keep arriving fast. That means both more opportunity and less time to prepare for each wave. The safety camp’s call for slower, more tested releases would actually give Pakistan more breathing room to adapt. You can explore how Pakistani institutions are getting ready for this in our earlier coverage of AI slowdown calls and what they mean for Pakistan and the broader innovation ecosystem discussed in our report on the Quantum Valley Pakistan launch.
What Comes Next
The argument is not over. Trump has no plans to change course, but the CEOs who called for a slowdown are not backing down either. Amodei proposed independent third-party evaluators embedded at the labs, coordination among democratic governments on safety standards, and eventually a global effort that would include authoritarian countries. Anthropic is committing to outside evaluators on its own, without waiting for others.
That means the debate will keep moving. Governments that have AI policies, like Pakistan, will face real choices about which framework to align with, the US go-fast model, the EU’s structured rules approach, or a third path suited to developing economies that want both growth and safety.
Frequently Asked Questions
Why did Trump reject the AI slowdown call?
Trump’s main reason was competition with China. He argued that slowing down would let China gain ground in AI, and that the US must stay ahead. He said some guardrails are possible but dismissed the safety warnings as exaggerated.
Which AI CEOs called for a slowdown?
The push followed an unprecedented joint effort by Anthropic, OpenAI and xAI to pace the industry amid growing safety risks. Elon Musk and Microsoft’s Satya Nadella also publicly backed the call, making it one of the broadest moments of agreement across rival AI companies.
Does Pakistan have an official position on AI safety?
The Islamabad AI Declaration, adopted in February 2026, is a strategic and diplomatic commitment that signals Pakistan’s intent to develop sovereign AI capabilities and to participate in international AI governance forums. Pakistan’s own policies stress explainability, human oversight and responsible use, which aligns more with the safety camp.
How could the US decision affect Pakistani users and developers?
If the US keeps developing AI at full speed, new tools, models and platforms will arrive faster. That gives Pakistani developers more options sooner, but also less time to adapt, train staff or update regulations. A global slowdown would slow that pace but also create more stability for businesses and regulators trying to keep up.
