Fear of AI Is Being Weaponized to Defend Tech Monopolies

As artificial intelligence advances at ever greater speed, it is posing major challenges to society. Many companies have begun laying off workers in roles that AI models can perform. The expansion of data centers is driving up energy consumption and the carbon footprint. Bad actors exploit AI to automate the spread of disinformation. Algorithms display bias, raising concerns about misuse and discrimination. AI is also being used to identify targets in war, leading to greater loss of life.

These are complex problems that demand serious attention from regulators. Yet mainstream AI policy discussions today rarely start from these immediate challenges.

Instead, global debate is swept up in prophecies of doom. This extinction narrative is pushed by some within the AI industry itself and echoed at the highest levels of decision-making: from legislators such as Senator Bernie Sanders, who has backed a ban on superintelligence, to warnings from the Vatican, or former U.S. Secretary of State Hillary Clinton when she called open-source AI a real threat.

It is entirely reasonable for world leaders to take seriously warnings amplified by the industry. But this chorus of influential voices inadvertently pulls our attention away from the many crises AI has already caused, affecting labor, the environment and the media.

More seriously, public panic is being turned into a tool for imposing a new kind of digital colonization – the concentration of power in the hands of a few AI corporations in a handful of countries.

The doomsday narrative provides a perfect justification for regulations aimed at open-source AI models – which are free or cheaper and allow users to modify them as needed. The argument put to legislators is simple and appealing: if advanced AI is a potential weapon of mass destruction, then only a small, trusted group should be allowed to develop it.

Under the banner of "safety," tech giants are seeking to write the rules to protect their monopoly position by blocking competition from smaller, cheaper open-source AI models.

This directly mirrors colonial tactics in history: preventing local industry from developing so as to force complete dependence on goods from the "empire."

Today the whole world depends on frontier AI systems developed by tech giants, especially the Global South – a region lacking the resources to build AI models from scratch, with compute costs now reaching more than $100 million.

The problem with using dominant AI models lies not only in their high cost but also in the fact that they often fail in local contexts. Take Kenya's recent rollout of its AI-run Social Health Authority (SHA) system. Because the imported model was built on assumptions that did not reflect Kenya's large informal economy, the system failed badly, overestimating incomes and overcharging the poorest citizens for health insurance.

To be effective, frontier AI systems must be tuned to local data. Since building from scratch is financially impossible for most countries, open-source AI is the only viable bridge for local innovators. More importantly, preserving open-source AI allows the Global South to step off the treadmill of Western human-simulating chatbots.

Developing countries do not necessarily need conversational AI systems; they need AI that helps solve development challenges in areas such as energy, health, agriculture, manufacturing and climate change mitigation.

If we crush open-source AI out of excessive fear of AI agents, developing countries will lose the main mechanism for building sovereign technology ecosystems that serve development.

Moreover, AI monopolies lock users into a cycle of data extraction and serious cognitive dependency. We have already seen the early stages of cognitive decline as users increasingly hand over critical thinking, creativity and decision-making to frontier AI systems.

This dependence creates a feedback loop in which growing reliance drives more data extraction, which in turn builds stronger systems, stripping people of autonomy and concentrating absolute control.

While the Global South is especially vulnerable because of historical inequality, this dependence traps anyone not at the frontier. Even wealthy countries in the Global North face the same threat of being locked in and exploited.

So what is the solution? Calling for an AI ban is certainly not it. Abandoning AI means giving up a technology with a special role in accelerating the discovery of new drugs and solutions to energy and climate crises. There is already evidence of this in the development of tools such as AlphaFold and MOFSeq.

Banning open-source AI models is also a poor option. Safety does not negate the right to equal access to AI. We can, and must, invest in robust safety research and rigorous testing for immediate harms. But we cannot let fear of hypothetical futures justify deliberately entrenching power in the present.

Protecting the future of AI means ensuring its development is decentralized, transparent and globally distributed. It means protecting open-source innovation as a public good and a basic weapon against digital colonization. It means proactively addressing the cognitive burden technology places on people, ensuring AI serves to augment human intelligence rather than erode it. It means anchoring ultimate accountability in humans, rather than anthropomorphizing AI agents that are not human.

The views in this article are the author's and do not necessarily reflect the editorial stance of Al Jazeera.