History offers a simple and enduring truth: mankind should fear not the tool it has created, but who wields it, with what motives, and against whom. From the stone axe to firearms, from the printing press to nuclear energy, at every turning point in the history of technology, the distinction between the medium’s essence and the intentions of the power instrumentalising it has determined the fate of civilisations. Yet, in the existential storms currently raging around artificial intelligence, this ancient truth is being deliberately obscured.
This analysis argues that the existential fear surrounding AI is not a genuine reckoning with technological risk, but a strategic narrative deployed by the very corporations that profit from AI’s harms. By framing the technology itself as the threat, these firms obscure the political economy of AI development, namely concentration of ownership, extraction of data, and deregulation of accountability, and position themselves as both the source of the problem and the only plausible solution. The real choice is not between embracing or rejecting AI, but between two models of governance: one that treats AI as private property to be exploited for profit, and one that treats it as public infrastructure to be governed in the common interest.
The Betrayed Promise of the Internet and Digital Exploitation
One need not look far into the past to understand this picture. Let us recall the optimistic expectations voiced during the early days of internet technology: a new, free, and inclusive public sphere that transcended the boundaries of time and space, democratising access to information. This vision has foundered today not because of the technical inadequacy of internet technology, nor because it has an inherently “dangerous” nature. The problem arose from the operational logic of digital platforms and their greed for exploitation.
Like any tool bearing the potential for utility, when social media was instrumentalised by the capitalist profit motive, the very platforms bearing the name “social” devolved into apparatuses of isolation, anger, and polarisation that transformed societies. This algorithmic architecture, which fractures attention spans and drives the general public—especially the youth—into emotional crises, has rewarded amplified fallacies over the dissemination of truth and context. Indeed, recent documentation of Meta knowingly concealing the harms on its platforms and consequently facing colossal settlements has shown, with concrete evidence, that this exploitative mindset is not an isolated accident but a systemic choice. Today, this very same mindset has taken the steering wheel of artificial intelligence, as if nothing had happened.
The Political Economy of Existential Fear: Fearmongering and the Evasion of Responsibility
The recent outcry from tech-giant executives— “Artificial intelligence could spell the end of humanity”—might, at first glance, seem like a reckoning of conscience.
However, this narrative of impending catastrophe can arguably be interpreted as serving two highly rational and calculated strategic objectives:
- The Evasion of Responsibility and De-subjectification: Instead of debating the copyright infringements, data theft, algorithmic biases, and labour exploitation engendered by technology, propagating “doomsday scenarios” obliterates existing legal and ethical responsibilities. By promoting the narrative that algorithms act autonomously and have spiralled out of control, it conveniently pins accountability for profit-driven decisions made by human hands on an abstract technology.
- Barriers to Market Entry and the Commercialisation of Security: By presenting a perception of a threat to mankind, the message is that AI is exceedingly dangerous and only incumbent tech giants can safely develop and regulate it. By leveraging this climate of fear to advocate for prohibitive regulations and steep compliance costs, there appears to be a concerted effort to erect formidable barriers to market entry. This strategy likely risks marginalising emerging rivals, open-source initiatives, and other nations, making it exceedingly difficult for them to participate in the technological race. The irony is that once this manufactured panic serves its purpose, these same corporations will pivot to marketing their technologies as ‘safe’—provided users pay for new premium versions. Even to be protected from this perceived danger, we will paradoxically be expected to pay them subscription fees, stand idly by as they commodify our data, and ultimately remain trapped under the digital hegemony of these very same global corporations.
Data Sovereignty and a De-Colonial Alternative
If artificial intelligence is to spell disaster for humanity, the means to prevent it cannot be found through a perspective trapped within the dichotomy of either complete capitulation to technology or its outright rejection. What is required is for states and societies without a Western—that is, a colonial and hegemonic—mindset to develop artificial intelligence models vastly superior to existing ones, driven by ethics and focused on the public good.
This transformation starts with data sovereignty. States and societies must treat their own data not merely as digital refuse, but as “sovereign soil”. Without losing momentum, every nation with the means must build its own models using its indigenous resources—models that are far more effective, sustainable, and less energy-intensive.
Yet, the path to data sovereignty is not only a matter of political will. It requires confronting four structural obstacles that the current debate often obscures. First, compute infrastructure: advanced AI models require vast computational resources, most of which are concentrated in the hands of a few American and Chinese firms. Without access to semiconductor supply chains, cloud infrastructure, and data centres, sovereignty remains aspirational. Second, talent: the expertise required to build frontier models is scarce, highly mobile, and currently concentrated in the same corporate ecosystems that dominate the market. Third, capital: the investment required to train and maintain competitive models runs into the billions, raising the question of whether state-led alternatives can achieve viability without becoming dependent on private capital. Fourth, governance: state-led AI could reproduce the same extractive logics it claims to oppose.
Therefore, the de-colonial alternative is not simply about building better models. It is about building different institutions that are public, accountable, and transparent, which ensure AI serves citizens rather than the state or the market.
Algorithms nourished by accurate information and operating not for profit maximisation but for the “common good” can counteract the devastation wrought by current social media platforms. Domestic and public artificial intelligence frameworks that present information to users without stripping it of context, and that regard individuals not as “consumers” to be exploited but as “citizens” with inherent rights, will reclaim digital power for society.
Beyond the Technological Race: An Ethical and Political Struggle
Ultimately, the artificial intelligence debate is not merely a technical coding race. It is a struggle over who will write the rules of the digital future, and, more fundamentally, over whether that future will be governed by the logic of extraction or the logic of the commons.
Technology becomes a catastrophe not because of the algorithm itself, but because of the exploitative mindset that governs it. Confronting this mindset requires three commitments: reclaiming data as sovereign infrastructure, building public institutions capable of governing AI in the common interest, and refusing the false choice between corporate control and technological rejection.
The choice, then, is not between embracing AI and fearing it, but between a democratic public-sphere-led digital order and a techno-feudal one.
