Concerns over the risks posed by superintelligent artificial intelligence (AI) have intensified amid calls for increased regulation and oversight of advanced AI technologies. Elon Musk, CEO of SpaceX, characterized recent demands by leading AI firms for regulatory measures as a potential strategic move to limit competition, describing it as a "psy op" aimed at restricting progress on cutting-edge AI development.
This skepticism was echoed by U.S. Vice President JD Vance, who expressed unease at the prospect of frontier AI companies appealing to the government for tighter controls, likening it to a "Trojan horse." Critics argue that such regulation could lead to "regulatory capture," favoring established players and hampering startups and emerging competitors.
Conversely, proponents of AI governance emphasize the importance of timely regulation to prevent catastrophic risks. Dario Amodei, CEO of AI developer Anthropic, advocates for policy frameworks while the industry remains concentrated among a few key actors to reduce the chances of accidents and uncontrolled technology proliferation. Supporting this view, computer scientist Yoshua Bengio, a pioneer in AI research, dismissed claims of regulatory posturing, noting that strict controls would likely impose significant financial costs on companies preparing for major stock market listings.
Experts warn that the potential development of superintelligent AI could represent a danger surpassing even nuclear weapons. Former UK Defence Secretary Des Browne highlighted this unprecedented risk, a point endorsed by Toby Ord, senior researcher at Oxford University’s AI governance initiative, who cautioned about the irreversible nature of such breakthroughs once achieved. However, the emergence of superintelligent AI remains speculative, with forecasts from the AI Futures Project placing it no earlier than late 2028, and OpenAI’s CEO Sam Altman suggesting a timeline extending to 2035.
Though AI is not currently integrated into nuclear command and control systems, the Rand Corporation has studied possible risks related to AI and nuclear weapons. Its findings suggest safeguards prevent AI from triggering nuclear launches, assuming human oversight remains intact. The United Nations has reinforced this position, reaffirming that human control should be maintained over nuclear arsenals. Nonetheless, AI could exacerbate nuclear tensions through disinformation campaigns or cyberattacks influencing decision-making, according to the International Campaign to Abolish Nuclear Weapons.
U.S. Senator Bernie Sanders has called for a moratorium on superintelligence research, urging a pause in advanced AI development due to the dangers signaled by industry insiders. In stark contrast, former President Donald Trump dismissed calls to slow AI progress, framing such advocacy as part of a "sick conspiracy" and asserting that effective leadership alone constitutes sufficient regulation.
The debate between accelerating and decelerating AI innovation also reflects broader geopolitical anxieties, particularly regarding competition between the U.S. and China, the two leading nations in AI research. U.S. Treasury Secretary Scot Bessent emphasized the strategic imperative to maintain technological superiority over China, warning that falling behind could have profound consequences. This dynamic fuels a perspective likening AI to a new form of strategic weaponry, reminiscent of Cold War nuclear arms races.
Chinese officials have rejected this framing, cautioning that adversarial rhetoric and confrontational policies hinder global governance efforts around AI. Some analysts suggest that calls for slower development by U.S.-based companies such as Anthropic may partly aim to curtail the rise of open-source Chinese AI models, which are both cost-effective and competitive but threaten the profitability of leading Western AI firms.
Industry voices in favor of continued research, like Pedro Domingos of the University of Washington, argue that greater understanding and control over AI will result only from sustained investigation rather than restrictions. Palantir’s CEO Alex Karp echoed this view, noting that in the face of international rivals, pausing AI development is not a viable option.
The ongoing debate underscores the complex interplay of technological risk, economic interest, and geopolitical rivalry shaping the future trajectory of artificial intelligence.