The Painful Dentist Visit That Big AI Predicted

In 2023, OpenAI CEO Sam Altman sat before Congress and delivered a warning that seemed almost theatrical: if AI goes wrong, it can go quite wrong. He asked for government help to prevent catastrophe. Dario Amodei, CEO of Anthropic, wrote in his essay “The Adolescence of Technology” that the coming years would be “impossibly hard” and would test humanity’s character. Both executives were essentially telling the public: this is going to hurt, and we don't want to be blamed when it does.

Now, three years later, that pain has arrived. A Politico report reveals that Big AI wants the Trump Administration to speak clearly about what is not okay. The same companies that once lobbied against regulation are now quietly grateful that the government has finally taken a hard stance—even if that stance has come with abruptness, opacity, and strictness that frustrates them. Dean Ball, recently hired at OpenAI as Head of Strategic Futures, summed it up: “I’m glad they’ve arrived to the conclusion that they have—to take this stuff seriously.”

The metaphor of a dentist showing the patient the tools before extraction is apt. Altman and Amodei showed America the terrifying tray: the potential for catastrophic failure, the need for guardrails, the impossibility of a smooth ride. But unlike a dentist who is invited, Big AI arrived unbidden. They had no certifications, yet they made huge promises about transforming society. And now the patient—the American public—is not happy. According to an Anthropic survey, only 15% of Americans trust AI companies to make decisions about development. 7 in 10 oppose data centers in their neighborhoods. And 87% believe foreign governments will use AI to attack the U.S. within 20 years.

The Trump Administration’s Two-Faced Approach

When President Trump and Vice President Vance took office, they signaled an aggressive pro-AI stance. In a February 2025 speech in Paris, Vance declared that regulation would paralyze one of the most promising technologies in generations. He essentially told the AI industry: go ahead, extract the teeth, no questions asked. The administration’s only major brush with regulation before this month was declaring Anthropic a supply chain risk—not because its models were dangerous, but because the administration wanted to control the most lethal tools in the world.

But then came the abrupt crackdown. Earlier this month, the Trump Administration issued an executive order requesting that AI companies submit their models for federal vetting. Then it went further: it suspended Anthropic’s Claude Fable 5 model, effectively imposing a moratorium on new releases. The move shocked the industry. An anonymous policy advisor for frontier AI companies told Politico, “It feels like they’re walking on eggshells a little bit.” OpenAI, while publicly cooperating, expressed frustration in a blog post: “We don’t believe this kind of government access process should become the long-term default.”

The result is a confusing regulatory landscape. Laws governing AI’s capabilities and role in society remain absent. What is allowed depends on whether Donald Trump is pleased. He reportedly dislikes guardrails that can be jailbroken, as was alleged with Fable 5. He also dislikes China-linked groups gaining access to models intended only for VIPs. So the current approach is less about structured policy and more about executive whim.

The Global Race and the Cost of Uncertainty

Meanwhile, rival labs in China are pushing ahead. Cybersecurity experts warned that the Fable/Mythos ban could allow Chinese AI developers to seize the opportunity while U.S. labs get bogged down figuring out what is allowed. The Trump Administration’s crackdown, although welcomed by some safety advocates, risks handing the lead to competitors. The U.S. regulation plan so far has circumvented Congress, relying on executive orders that can change with the political wind.

Big AI’s earlier warnings now seem prescient. Altman and Amodei were not being disingenuous; they were trying to set expectations. They knew that the technology would create immense pressure, and they wanted the government to share the burden of responsibility. But the government, under Trump, first refused to regulate, then applied a sledgehammer. The industry is now in a holding pattern: companies like OpenAI are releasing new models (the GPT 5.6 series) only to small groups of VIP customers while working with the administration to develop “a repeatable process for future model releases.”

Public Distrust and the Missing Conversation

The deeper issue is public distrust. Americans are pessimistic about AI. They see the benefits but fear the risks. The fact that Anthropic itself conducted a survey showing only 15% trust is a damning indicator. People want slower development, clearer rules, and real oversight. Yet the conversation in Washington remains dominated by industry lobbying and executive branch power struggles. Congress has not passed meaningful AI legislation. The debate is trapped between two extremes: the pro-innovation, no-regulation camp and the sudden crackdown camp. There is no middle ground where safety and innovation are balanced.

As Saif Khan, a former Biden Administration tech advisor, noted, the administration’s actions have resulted in an almost complete moratorium on new releases, which is starting to seriously impact companies’ bottom lines. But the pause might be temporary. Politico reports that Claude Fable 5 could be turned back on in a matter of days. The pattern is clear: the White House will lift the ban, allow releases to resume, then crack down again when something displeases them. This uncertainty is bad for investment, bad for safety, and bad for global leadership.

In the end, Big AI had a point. They told us the tools were dangerous. They showed us the tray. They asked for clear rules. And now we are living in the painful future they warned about, with a president who has paused the procedure but plans to restart it without changing anything fundamental. The public remains anxious, the industry is frustrated, and the world watches as the U.S. fumbles its way through the most consequential technological transition since the internet.

Source: Gizmodo News