A new United Nations report warns that the development of artificial intelligence may exacerbate global inequality and proposes a shared framework for how to responsibly develop AI, as adoption and investment into the technology accelerates unevenly across the world.
“The more AI advances without shared rules, the less say governments and people will have in the outcome,” said António Guterres, the UN secretary general, at a press conference on Wednesday. “Our message to governments is simple: do not wait … the science is here. We can no longer say we did not know what we do.”
The sweeping analysis from the independent international scientific panel on AI, established by the UN general assembly last year as “the first global scientific body on AI”, details AI’s risks and opportunities – from transformative capabilities in agriculture and education, to catastrophic outcomes when bad actors deploy AI to commit fraud and influence elections.
“Access to AI tools alone does not produce equal benefit,” the report states. “Countries that rely on foreign models, cloud infrastructure and data pipelines may gain access to AI while losing practical control over its standards, safeguards and local fit.”
At the press conference, co-chair of the panel, journalist Maria Ressa stressed that AI’s “pace is not slowing, the power is concentrating, and control is not guaranteed”. The report dropped one week before the UN hosts the inaugural global dialogue on AI governance for governments and experts.
The preliminary report also functions as a toolkit, offering initial, broad guidance to UN member states on ways to capitalize on AI’s potential for growth across industries, while minimizing and addressing threats. Suggestions include developing local AI infrastructure, such as datacenters, improving AI literacy in schools and the workforce, investing in developers, building AI safety institutes, creating strategies to combat disinformation and continuously measuring how AI systems behave after release, “with real users, real tasks and real environments”.
While more than a billion people now use AI weekly, access and types of usage vary widely across the world, “with adoption across the global south lagging far behind the global north”, the report states. The US and China dominate in the development of leading AI models, as well as investment into compute infrastructure, which encompasses the hardware, memory, networking and storage required to run powerful AI models.
“The concentration of AI capabilities in a small number of firms and countries could enable authoritarian capture and undermine democratic accountability,” the report states.
The panel advises countries lagging behind in AI development to consider significant investment in computing and data infrastructure. Attracting this money requires securing a reliable energy supply and building datacenters, they note. The report does, however, acknowledge the environmental costs of datacenters, including their large energy and water consumption, and potential for greenhouse gas emissions.
The authors also describe challenges in evaluating safety and providing oversight of increasingly powerful AI models.
“Most countries, including many advanced economies, lack the technical expertise to assess the most capable ‘frontier’ models or to participate meaningfully in their governance,” they write.
The panel of 40 independent scientific experts from across the world stated that this report is “the first of its kind”. The UN, they argue, “is the foremost global forum on transboundary risks of this scale” – and its approach is “scientific, not political”.
Differences in language and internet access compound the digital divide.
after newsletter promotion
“Artificial intelligence leaves most languages behind,” the report notes. While generative AI tools perform well in English and other widely used languages, “most languages are either excluded or have much lower performance”.
These disparities can have significant implications, particularly in a healthcare context. The report points out an example of a machine translation of Tigrinya mixing up smallpox with syphilis, gonorrhoea as diabetes, and the phrase “you have been given intravenous antibiotics” as “you have been given intravenous insecticides.”
“These mistranslations can be life-threatening,” the report notes.
Some regions lack stable internet access, let alone adoption of AI models. More than 2 billion people – almost a third of the world’s global population – are completely offline, according to the International Telecommunication Union.
At Wednesday’s press conference, journalists pressed the independent panel on why they didn’t apply more pressure to secure commitments from countries and recommend an international body to analyze powerful AI models before they are released to the public.
The panel pushed back, and described the report as focused on giving facts, not prescribing policy. If they leaned too heavily on recommendations they feared their work would become politicized and it would “pollute (their) ability to provide scientific evidence”, Ressa said.
“It may sound like a cop out that we’re not making policy recommendations, but frankly, when you have the scientists together in the room, they tell you what (they) know,” Ressa said. “That’s why it is usable in Washington and Beijing and Manila. The prescribing happens next week in Geneva … where the states sit at the table.”
