Debate has intensified in scientific, media and political circles worldwide over the extraordinary pace of artificial intelligence development, the fear of losing control over it, and the prospect that humanity is entering a new era defined by an unprecedented challenge: AI systems with potentially unlimited capabilities that could govern human life with little or no human input. Alternatively, critics warn of the opposite danger, that colonial powers, certain lobbying groups and a handful of individuals could tighten their grip on human life by controlling AI.
اضافة اعلان
The Stanford AI Index 2025 shows a marked rise in recorded cases linking AI use to real harm or serious problems: 233 cases in 2024, compared with 149 in 2023, an increase of roughly 56.4 per cent in a single year. These cases include the use of deepfakes for disinformation or fraud, flawed decisions produced by algorithmic systems, data leaks, and financial, social or security damage caused by AI systems or their misuse. The figure does not capture the full picture, however: the underlying database relies heavily on incidents that were publicly reported or covered by the media, and other cases may remain undisclosed or undocumented.
These numbers do not mean that AI is killing people on a wide scale. They do reveal an important trend: as intelligent systems spread into more sensitive domains, the opportunities for error and misuse multiply.
This article will not examine the capacity of a small group of people, whose views may clash with those of the majority, to shape how individuals think through algorithms or tailored responses that steer users toward particular conclusions. That, in the author's view and in the view of many others, is already a reality. There are, however, larger dangers to consider.
Imagine an AI system making a mistake in a decision that affects a person's life. What do we do then? In this context, the International AI Safety Report 2026, prepared by more than one hundred experts with support from more than thirty countries and international organisations, notes that current systems can produce fabricated information, defective code or misleading medical advice. Such errors can cause physical, psychological or financial harm, particularly when users rely on them without any human review.
The danger grows when AI moves from offering a suggestion to taking action itself. If a system can operate equipment, control a vehicle, manage infrastructure or execute financial transactions, a mistake is no longer a wrong answer on a screen; it becomes a physical event with real consequences for large numbers of people.
This risk is acute in medical diagnosis, radiological analysis and drug discovery. The danger surfaces when people treat an AI system as an infallible doctor, especially as users increasingly defer to applications, assuming the output of an intelligent system is superior to human judgment. The consequences of a wrongly prescribed drug or a misdiagnosis that costs someone their life are not difficult to imagine.
The risks escalate further in military contexts. The Stockholm International Peace Research Institute (SIPRI) studied autonomous weapons systems and AI-based military decision-support systems in 2025 and warned of the dangers of reducing human involvement in decisions on the use of force, and of the potential for unintended harm. The implication is clear: as reliance on such systems grows, an error or a breach could trigger dangerous weapons, ignite a conflict or, in the worst case, destroy cities.
AI systems could also provide sensitive information to untrustworthy actors, accelerating their ability to develop dangerous weapons. The threat from cybercrime and cyberattacks is no less serious. AI can write software and identify vulnerabilities, a capability that is useful to defenders of electronic systems but equally useful to attackers.
The International AI Safety Report 2026 states that AI systems have become capable of detecting software vulnerabilities and writing malicious code, and that criminal groups and state-linked actors are already incorporating AI models into some of their operations. In one competition cited by the report, an AI agent identified 77 per cent of the vulnerabilities present in real software within the test environment used. Applied to attacks on hospitals, power grids, water systems, air transport or emergency services, the scale of that threat becomes apparent.
The same technology underpins deepfakes, which can erode public trust in governments, thinkers, values and beliefs, enable character assassinations and fabricate video footage of real individuals. Beyond this, the concentration of AI power in the hands of a few raises concerns about control over labour markets, growing violations of workers' rights and harm to the most vulnerable.
The author is not opposed to the development and progress of AI, but objects to monopolisation and a form of colonial control that has shifted from one state dominating another to a small group of individuals dominating the entire world. The concern is not with scientific progress as such, but with uncalculated risks that threaten human life, and with the concentration of this power in the hands of a limited number of companies or states, which could produce an unprecedented imbalance in the distribution of economic, political and knowledge-based power globally.
AI development must therefore be governed by strict rules that go beyond traditional frameworks allowing powerful states to act freely while restricting less powerful ones. This concentration of power must be recognised as a genuine danger that requires a collective response. Less developed countries, for their part, should understand that scientific advancement and balance with major powers is the surest path to a future built on pluralism and multipolarity, rather than leaving the field to a global order that has demonstrated, on more than one occasion, that human justice is not central to its thinking or its actions.
Artificial intelligence is a great gift that could, in an instant, become an uncontrolled instrument that destroys what is most valuable in human life and leaves the future filled with fear.
This article was originally written in Arabic by Al Ghad for Al Ghad. It was translated into English with AI assistance and edited by Jordan News.
