Al Jazeera Opinion Warns AI Safety Fears Could Strengthen Tech Monopolies

An Al Jazeera opinion article has argued that public fears about catastrophic artificial intelligence risks are being used to support policies that could restrict open-source systems and strengthen the position of major technology companies.

The author says the debate should give greater attention to harms already associated with AI, including worker layoffs, rising energy use and carbon emissions from data centres, automated disinformation, algorithmic bias and the use of AI to identify targets in warfare.

According to the article, public and policy discussions are frequently being redirected toward hypothetical apocalyptic scenarios instead of focusing on immediate effects involving employment, the environment and communications.

Concerns over concentration of power

The author argues that warnings about existential AI risks are helping to concentrate development and control in the hands of a small number of corporations and countries. The article describes this process as “a modern regime of digital colonisation” and says public panic is being used as a political tool.

It further claims that technology companies are seeking safety rules that could limit competition from smaller and cheaper open-source models. Such systems are generally described as more affordable or freely available and can be modified by users.

“Under the banner of ‘safety’, tech giants are attempting to write the rules to protect their monopolies by shutting down competition from smaller, cheaper open-source AI models,” the author writes.

The article does not identify specific companies or proposed laws. Its claim that industry participants are seeking restrictions is presented as the author’s analysis.

Implications for the Global South

The opinion piece says countries in the Global South are particularly dependent on frontier AI systems developed by large technology companies. It argues that this dependence can lead to user lock-in, further data extraction and increased control by the companies operating the systems.

Kenya is presented as an example of the risks associated with importing AI systems without sufficient attention to local conditions. The author says an AI-driven Social Health Authority system allegedly overestimated incomes and charged poorer citizens more for healthcare premiums because it did not accurately reflect the country’s informal economy.

Those claims about Kenya are made by the article’s author and are not independently confirmed in the supplied material.

The article says developing countries need AI applications in areas such as energy, healthcare, agriculture, manufacturing and climate mitigation. It argues that many local innovators cannot afford to build frontier models from the beginning, with the cost of computing alone described as potentially reaching upwards of $100m.

For that reason, the author calls open-source AI “the only viable bridge for local innovators”, saying it can allow developers to adapt systems to local data and circumstances without creating frontier models from scratch.

Call for safety and access

The author rejects calls to ban AI or open-source AI, arguing instead for stronger safety research, rigorous testing and broader access. The article also supports a more decentralised, transparent and globally distributed approach to AI development.

It cites tools associated with drug discovery and energy and climate challenges as examples of potential benefits from AI. At the same time, it calls for safeguards against harmful uses, attention to possible cognitive effects and continued human accountability.

“Safety does not negate equitable access to AI,” the author writes, maintaining that safety measures should not be used to prevent local innovation or deepen dependence on a limited number of technology companies.