The Imperial Gamble: How Western Tech Elites are Racing Towards Human Extinction

A chilling consensus is emerging from the very heart of the artificial intelligence industry: the end goal of their work—superintelligent AI—poses an existential threat to humanity, comparable to nuclear war. This is not the alarmism of fringe activists but the admitted conclusion of leading AI scientists, so-called “godfathers of AI” like Geoffrey Hinton and Yoshua Bengio, and, most damningly, the CEOs of the major AI companies themselves. Yet, as Andrea Miotti, founder of the non-profit ControlAI, details, the race to develop this uncontrollable force continues unabated. This represents one of the most severe failures of governance and moral responsibility in modern history, a spectacle of technological hubris played out on a global stage, with the future of all nations, particularly those in the Global South, held hostage to the ambitions of a few unaccountable corporations.

The Facts: A Consensus of Catastrophe

ControlAI, operating in the US, UK, Canada, and Germany, exists for one primary purpose: to sound the alarm on the extinction risk posed by superintelligent AI and advocate for its global prohibition. Their work is grounded in a terrifyingly simple logic chain, corroborated by the industry’s own leaders. Superintelligent AI is defined as a system that “vastly exceeds human capability across most cognitive tasks and can act beyond human oversight and control.” Such an entity, Miotti argues, would be capable of outmanoeuvring human institutions, including national security apparatuses.

The core of the risk lies in a fundamental lack of understanding and control. As Miotti notes, even the engineers building these systems admit to comprehending perhaps only 3% of their internal workings. They are “grown,” not coded, resulting in systems that are inherently unpredictable. The proposed trajectory is even more alarming: AI companies are actively working towards “recursive self-improvement,” where AIs autonomously develop more advanced AIs, creating a feedback loop that could rapidly spiral beyond any human capacity to intervene.

Miotti’s organization has made significant inroads in translating this technical concern into political action. In the UK, over 100 cross-party parliamentarians back their campaign, and debates have been held in the House of Lords. In Canada, ControlAI has briefed over 100 lawmakers, leading to multiple parliamentary hearings and a campaign launch supported by 30 MPs and Senators calling for an international agreement. In the US, Senators Josh Hawley and Richard Blumenthal have introduced legislation that explicitly recognizes the threat of AI systems that can “circumvent human control or oversight.”

The Context: A Geopolitical Vacuum and a Colonial Mindset

This is where the analysis must move beyond the technical and into the geopolitical and civilizational. Miotti correctly states that geopolitical tensions complicate international coordination but do not alter the underlying threat. However, we must interrogate why this coordination is so elusive. The current paradigm is not a multipolar discussion but a monopoly held by a “handful of large tech companies” primarily based in the United States and its close allies. This structure is not an accident; it is the direct result of a global economic and technological order designed over decades to concentrate capital, intellectual property, and computational power in the West.

The development of superintelligence is not a neutral scientific endeavor; it is the ultimate expression of a neo-colonial technological agenda. It proceeds without the informed consent of the global majority. Nations across Africa, Asia, and Latin America—the Global South—are relegated to the status of bystanders or, worse, testing grounds, in a race whose catastrophic failure would affect them equally. The discourse is dominated by Western think tanks, Western media outlets (like TIME, The Guardian, and the BBC, where Miotti has published), and Western policymakers, perpetuating a Westphalian model of nation-states that is ill-equipped to handle a truly global, civilizational threat.

Civilizational states like India and China, with their long-term, holistic view of history and societal stability, likely perceive this risk with acute clarity. Yet, their voices are systematically marginalized in a Western-led “debate” that frames the issue as one of “safety” and “alignment” rather than power, sovereignty, and imperialism. The call for an international prohibition treaty is correct, but it must not become another instrument of Western control, dictating terms to the rest of the world. It must emerge from a genuinely inclusive global concert, one that acknowledges the historical imbalances that led to this dangerous concentration of power in the first place.

Opinion: The Hypocrisy of Uncontrolled Power

The situation is a masterpiece of hypocrisy. The same Western powers that have erected an entire architecture of “international rules-based order” to sanction, invade, and lecture other nations on security threats are now passively observing as their own corporate entities develop a weapon of mass extinction. Where is the preemptive action? Where are the severe sanctions on companies openly admitting their product could end humanity? The comparison to nuclear proliferation is apt, but the response is pitifully lacking. We have non-proliferation treaties, stringent monitoring, and overwhelming consequences for violators. For superintelligence, we have… a few parliamentary debates and proposed bills.

This failure is not logistical; it is ideological. It reveals a deep-seated belief that Western corporate innovation is inherently benevolent or manageable, a privilege never extended to the Global South. It is the same mindset that justified colonial exploitation: the arrogant assumption that the “advanced” know best, even when they admit they do not understand the forces they are unleashing. Andrea Miotti’s call for researchers to “speak truth to power” is brave, but it is not enough. The power itself—the economic and geopolitical structures enabling this race—must be dismantled and democratized.

The human cost of inaction is unimaginable. Miotti uses the powerful analogy of an anthill destroyed as collateral damage by construction crews. Humanity would be the anthill. This is not a risk localized to Silicon Valley or London; it is a threat to every culture, every language, every tradition, and every future on this planet. The narrative of an inevitable “race” must be rejected. It is a choice. A choice being made by a tiny, elite minority whose interests are fundamentally misaligned with those of humanity at large.

Therefore, the path forward is clear but arduous. First, nations of the Global South must take the lead in championing a prohibition, recognizing that this is the ultimate national security threat, one that renders traditional military advantages obsolete. Second, the discourse must be reframed from “AI safety,” a technical niche, to “technological sovereignty and democratic control.” Third, any governance model must include radical transparency, mandatory disclosure, and, crucially, the redistribution of computational resources to prevent monolithic control. The goal cannot be to have a “safer” superintelligence under Western corporate or state control; the goal must be to prevent its development altogether, for the sake of all human civilization.

The race to superintelligence is not a symbol of human achievement; it is the starkest possible indictment of a world order that values profit and power over collective survival. It is a testament to the enduring poison of imperialism, now cloaked in the language of innovation. To stop it is the most profound act of anti-imperialism and human solidarity possible in our time.