“The more artificial intelligence advances without common rules, the less influence governments and people will have on the results,” United Nations Secretary-General António Guterres said during a press conference on Wednesday.
He added: “Our message to governments is clear: Don’t wait… the science is there. We can no longer claim that we did not know what we know.”
The report was issued by the Independent International Scientific Committee on Artificial Intelligence, which was established by the United Nations General Assembly last year, as the first global scientific body specialized in this file.
The report presents the opportunities and risks of artificial intelligence, from its potential uses in agriculture and education, to threats related to fraud and influencing elections when used by bad actors.
The report noted that “access to artificial intelligence tools alone does not achieve equal benefit,” explaining that countries that rely on foreign models and external cloud architectures may obtain the technology, but they practically lose control over its standards, guarantees, and suitability for the local context.
Journalist Maria Ressa, co-chair of the committee, said that the pace of development of artificial intelligence “is not slowing down, but rather power is concentrated in it, and control over it cannot be guaranteed.”
The report comes a week before the United Nations hosts the first global dialogue on artificial intelligence governance, with the participation of governments and experts.
The report suggests practical steps for countries, including developing local infrastructure for artificial intelligence, such as data centers, enhancing awareness of this technology in schools and the labor market, investing in developers, and establishing institutes specialized in artificial intelligence safety.
It also calls for developing strategies to combat disinformation, and measuring the behavior of artificial intelligence systems after their launch, with real users, tasks and environments.
Although more than one billion people use artificial intelligence weekly, the report indicates that its adoption in countries of the Global South is still far behind compared to countries of the Global North.
The report confirms that the United States and China dominate the development of leading models and investment in the computing infrastructure necessary to operate advanced artificial intelligence systems.
He warned that concentrating AI capabilities in a limited number of companies and countries could open the door to “authoritarian takeover” and undermine democratic accountability.
The committee called on countries lagging behind in this field to invest in computing and data infrastructure, while securing reliable energy sources and building data centers.
But the report also pointed out the environmental cost of these centers, due to their large consumption of energy and water and the possibility of increasing greenhouse gas emissions.
He also pointed out that most countries, including advanced economies, lack sufficient technical expertise to evaluate the most advanced artificial intelligence models or participate effectively in their management.
The report’s authors stressed that the United Nations represents the most prominent global forum for dealing with cross-border risks of this magnitude, stressing that their approach is “scientific, not political.”
