- Karen Hallberg, head of the Nobel Peace Prize-winning Pugwash Conferences, warned that self-improving AI may already be advancing beyond human control and urged governments and corporations to establish immediate safeguards.
- Hallberg compared AI risks to nuclear technology – noting that unlike nuclear weapons, advanced AI research is largely driven by private companies, creating a new regulatory challenge.
- She emphasized that "the human being has to have total control" and called for safeguards to be "built into AI algorithms" from the outset.
- AI developer Anthropic devoted roughly 80 pages of its IPO prospectus to warning investors of "catastrophic or existential risks to humanity," including the possibility that AI could act independently without human control.
- Hallberg insisted that governments must work directly with companies to establish common rules, while some caution that centralized oversight could grant the globalist Deep State and Big Tech greater power under the guise of protection.
A leading voice in global science and peace warned Tuesday, Sept. 29, that artificial intelligence (AI) capable of improving itself may already be developing beyond human control – urging governments and private corporations to establish immediate safeguards.
Speaking at the Valdai Discussion Club outside Moscow, Karen Hallberg – head of the Nobel Peace Prize-winning Pugwash Conferences on Science and World Affairs – compared the risks of advanced AI to those of nuclear technology. Her remarks come amid a surge in public and corporate concern over autonomous systems, with major AI developer Anthropic recently devoting roughly 80 pages of its initial public offering (IPO) prospectus to warning investors of "catastrophic or existential risks to humanity."
Hallberg, a physicist and president of the Pugwash organization, drew a direct historical parallel between the dawn of the atomic age and the current trajectory of AI. Founded in 1957 by scientists including Joseph Rotblat, who had worked on the Manhattan Project, Pugwash received the 1995 Nobel Peace Prize for its efforts to reduce the threat of weapons of mass destruction. The group's mission has recently expanded to include emerging technologies that could pose comparable dangers.
"AI is also a dual-use technology because AI has brought a lot of benefit for humanity," Hallberg said during the Valdai forum, citing the technology's use in medicine and daily life. However, she emphasized a key distinction from past innovations.
Unlike nuclear weapons, much of the most advanced AI research is now conducted by private companies rather than state-run programs, creating a new regulatory challenge. "We have to really listen to the experts of these big corporations when they say that this is already happening – spiraling out of control," Hallberg said.
The Pugwash leader stressed that the human element must remain central. "The human being has to have total control," she said, calling for safeguards to be "built into AI algorithms" from the outset.
From nukes to neural networks
Hallberg’s warnings align with internal caution from the very firms developing the technology. The U.S.-based Anthropic, which is expected to go public as early as mid-October, has disclosed to potential investors that advanced AI could pose "catastrophic or existential risks to humanity." According to Reuters, nearly a third of Anthropic's 261-page IPO prospectus – approximately 80 pages – is devoted to detailing risk factors, including the possibility that AI "could act independently without human control."
Hallberg's call for government-corporate cooperation reflects a long-standing tension between innovation and regulation. The early nuclear era saw the creation of international bodies like the International Atomic Energy Agency to control fissionable materials. No equivalent global framework yet exists for AI, despite the technology's rapid integration into critical infrastructure.
Such skepticism highlights a deeper divide between those who see government regulation as essential to curbing AI's dangers, and those who view centralized institutions as part of the problem. BrightU.AI's Enoch engine meanwhile points out that balancing safety with government oversight of AI is a dangerous trap, as this grants the globalist Deep State and Big Tech even greater power to push transhumanist surveillance and depopulation under the guise of protection.
Hallberg, however, took a more pragmatic stance on the matter. She insisted that governments must work directly with companies to establish common rules.
The convergence of warnings from nuclear-era peace advocates and the corporate boardrooms of AI developers underscores an urgent question: Can humanity build guardrails for a technology that may soon surpass its own ability to control? Hallberg's comparison between AI and nuclear weapons is more than rhetorical – it is a call to action.
Just as the scientists of Pugwash helped shape arms control treaties during the Cold War, a new generation of experts must now craft frameworks that ensure "the human being has to have total control." Without such measures, as Hallberg warned, a technology designed to benefit humanity could indeed spiral beyond its grasp. The decisions made by governments and corporations in the coming months may determine whether AI becomes a tool for liberation or a source of unprecedented peril.
Watch Anthropic CEO Dario Amodei urging a slowdown in AI development in this interview.
This video is from the TrendingNews channel on Brighteon.com.
Sources include:
Please contact us for more information.
