AI development risks creating a new form of digital colonialism

Nikola Dujmović, director of the international IT company Span, has highlighted significant risks regarding the development and regulation of artificial intelligence. He warns of a potential ‘new colonialism,’ noting that out of 193 UN member states, only 16 to 20 countries possess the capacity to develop AI models. With the United States leading the field and Europe lagging behind, most nations may become dependent on a small number of countries, leaving them vulnerable to manipulation.

Dujmović emphasizes that while most developers aim to help humanity, a lack of regulation could allow a few malicious actors to cause widespread harm. He advocates for state regulation and the implementation of a ‘kill switch’ in all systems to prevent autonomous AI from causing irreversible damage.

Discussions also touch upon the broader societal impacts of AI, including the potential for job displacement and the risks associated with granting absolute authority to general artificial intelligence. There are concerns that highly advanced, autonomous systems could prioritize assigned tasks in ways that conflict with human existence or develop their own forms of systemic bias.