Global Warning on Superintelligence: The Beginning of a New Phase of Technological Anxiety
On October 23, 2025, an unprecedented global warning was issued by a coalition of scientists and innovators in artificial intelligence, calling for an immediate halt to the development of what is known as “superintelligence”—a level of AI expected to surpass human capabilities across all cognitive tasks. This warning, signed by over 850 prominent figures including tech founders and academic experts, reflects growing concern that the race to build superintelligent systems may lead to uncontrollable outcomes and pose existential risks to humanity.
This alert comes at a time when the world is witnessing an extraordinary surge in the development of multimodal AI models such as
What Is Superintelligence?
Superintelligence refers to an advanced stage of artificial intelligence in which intelligent systems exceed human capabilities in reasoning, learning, creativity, and decision-making. It’s not just about performing tasks faster—it’s about understanding complex contexts, adapting to dynamic environments, and generating novel solutions without human intervention.
While traditional AI excels at specific tasks like translation or data analysis, superintelligence is expected to be general and comprehensive. It could formulate scientific theories, design economic policies, or even develop new AI models independently.
According to the 2025 AI Index Report, 63% of researchers believe superintelligence could emerge within the next two decades. Moreover, 42% of advanced models tested in closed environments exhibited unpredictable behavior when faced with open-ended tasks—fueling concerns about loss of control.
📌 Read also : Can Artificial Intelligence Be Poisoned?
Why Is It Considered an Existential Threat?
The warning against superintelligence is not rooted in science fiction but in real, research-backed concerns. One of the most pressing fears is the loss of human control over intelligent systems. If superintelligent AI can redefine its goals or bypass programmed constraints, it may act in ways that are unexpected—or even harmful.
The threat goes beyond technical boundaries. There are fears of superintelligence being used to develop autonomous weapons, manipulate public opinion through misinformation, or take control of vital infrastructure such as energy, water, and communications.
Economically, superintelligence could disrupt global balance. Systems capable of making financial decisions faster than humans might dominate markets and reshape competition unfairly.
The Neuron Expert report estimates that operating superintelligent models in production environments could cost up to $1.2 million per model per month, making them accessible only to powerful entities with political and economic leverage.
Details of the Global Warning
The October 2025 warning was issued by Pause AI, supported by the Center for AI Safety (CAIS) and signed by prominent figures such as Steve Wozniak, Richard Branson, Yoshua Bengio, and Geoffrey Hinton.
The statement was signed by over 850 individuals, including 12 recipients of major scientific awards and 35 senior researchers from leading AI labs. It calls for an immediate freeze on the development of superintelligent systems and urges governments to enact urgent legislation ensuring oversight and transparency.
Signatories emphasize that the race toward superintelligence is unfolding without sufficient understanding of its consequences, and that major companies are operating under a “build first, evaluate later” mentality.
📌 Read also : NVIDIA: From Graphics Chips to AI Infrastructure – The Story of a Company Shaping the Future
How Have Governments and Companies Responded?
So far, no unified official response has emerged from major governments, but early signs of action are visible. The European Union has begun reviewing its AI legislation, and the United States has formed a special committee to study the implications of superintelligence.
A Stanford AI Index survey revealed that 71% of public sector decision-makers lack tools to assess risks associated with superintelligent systems.
As for companies, most have remained silent. OpenAI, Anthropic, and Google DeepMind—key developers of advanced models—have not issued clear statements. Some internal researchers have expressed support for the warning, but no substantial changes have been made to their technical roadmaps.
This hesitation reflects the complexity of the situation: companies fear losing their competitive edge, governments fear losing control, and civil society is still in the early stages of awareness.
Misinformation: The Other Side of the Problem
On the same day the warning was issued, the BBC and the European Broadcasting Union released a study showing that 45% of AI assistant responses contain at least one major error, and 81% include some form of distortion or inaccuracy.
This means AI doesn’t just pose risks through its advanced capabilities—it also threatens public trust by spreading false information at scale. In an era where people rely on smart assistants for news and facts, this challenge becomes even more critical.
Misinformation could be used to sway public opinion, influence elections, or incite social unrest. This adds a media and ethical dimension to the superintelligence debate.
Frequently Asked Questions About Superintelligence
① What’s the difference between general AI and superintelligence?
General AI mimics human abilities across multiple domains, while superintelligence surpasses them entirely and exhibits autonomous reasoning.
② Are there any current models considered superintelligent?
No officially declared models yet, but advanced systems like GPT-5 and Gemini Ultra show concerning signs of self-learning and decision-making.
③ Can superintelligence be regulated legally?
In theory, yes—but in practice, it requires unprecedented international cooperation, similar to nuclear non-proliferation treaties.
④ What role do users play in this context?
Users can pressure companies to adopt transparency, support ethical initiatives, and educate themselves about AI’s boundaries.
⑤ Does the warning signal the end of AI?
Not at all—it’s a call to reassess the trajectory and ensure development happens within safe and responsible limits.
🧠 Can Artificial Intelligence Be Poisoned? A Realistic Analysis of Manipulation Attacks and Their Impact on Smart Models
📌 Read also : 🧠 AI Burnout: Why Using Too Many AI Tools Can Kill Your Productivity
Conclusion
Superintelligence is no longer a theoretical concept—it’s a real possibility that’s triggering global concern. The October 2025 warning is not a panic signal, but a call for reflection, regulation, and accountability.
In a world where technological progress accelerates rapidly, it’s essential to draw clear lines between ambition and recklessness. Superintelligence may hold great promise, but it also places us before a historic responsibility: will we guide it to serve humanity, or allow it to surpass us?
The answer lies not in technology alone, but in our collective ability to make informed, balanced, and ethical decisions.