AI’s ‘opportunity divide’ could deepen global inequality, UN official warns

Doha, Qatar – Artificial quality (AI) risks widening the spread betwixt affluent and mediocre countries, perchance deepening existing societal and economical inequalities, a elder United Nations authoritative has warned.

Developing countries look important barriers to benefitting from the AI revolution, Tomas Lamanauskas, the lawman secretary-general of the International Telecommunication Union (ITU), a UN bureau focused connected accusation and connection technologies, told Al Jazeera.

Recommended Stories

• list 1 of 3OpenAI says its agents turned disconnected ain safeguards
• list 2 of 3Trump says AI information centres should wage ‘dividends’ successful pre-midterm shift
• list 3 of 3Canadian steadfast drops writer Thelyson Orelien implicit AI usage successful novel

“There is simply a hazard … I deliberation the AI disagreement is truly the large interest due to the fact that it’s truly not conscionable astir the integer divide,” Lamanauskas said connected the sidelines of the UN Conference connected Trade and Development (UNCTAD) World Investment Forum successful Doha, Qatar connected Sunday.

“AI has the immense imaginable of powering health, education, nationalist services, industries, agriculture, trade. So a large disagreement could truly mean an accidental divide, and truly could deepen the socioeconomic divide,” helium said.

“So that’s truly the worrying part.”

Lamanauskas said the disparities are already evident successful integer infrastructure, citing the examples of Africa and Latin America, which unneurotic relationship for conscionable 3.3 percent of planetary information centre capacity.

“So you intelligibly spot inactive a alternatively large infrastructure divide,” helium said, adding that 2.2 cardinal radical worldwide stay without net access.

In a December 2025 report, the UN Development Programme (UNDP) cautioned that AI could trigger a caller epoch of inequality betwixt nations, reversing decades of advancement successful narrowing economical disparities betwixt nations.

The UNDP warned that countries with unreliable power, constricted connectivity, and a deficiency of technological skills hazard missing retired connected the economical gains of AI.

Bridging the integer infrastructure spread would necessitate an estimated $2.8 trillion successful concern by 2030, UN authoritative Lamanauskas said.

The disagreement betwixt nations is besides evident successful the adoption of AI, helium added. While astir 29 percent of working-age radical successful developed countries usage generative AI, that fig drops to 16 percent successful processing nations.

“This really affects your quality to beryllium productive and to beryllium portion of job-market opportunities,” Lamanauskas said.

Loss of jobs

Lamanauskas acknowledged increasing concerns astir AI-driven occupation losses, citing estimates that the exertion could displace 90 cardinal jobs, but besides make 170 cardinal caller ones.

However, helium warned that adjacent if determination is simply a “net affirmative impact”, galore radical volition personally acquisition disruption, arsenic an estimated 40 percent of workplace skills could alteration by 2030.

He besides pointed to connection inequalities, revealing that lone 7 percent of African languages are represented successful AI translation tools, limiting entree for millions of imaginable users.

According to the 2026 International AI Safety Report released successful February, astir 1 cardinal radical usage AI for a assortment of tasks, including seeking guidance connected aesculapian wellness and writing.

While nary factual information exists, the study estimated that astir 60 percent of jobs successful precocious economies and 40 percent successful emerging economies could beryllium exposed to AI-related impacts.

“The fears are real,” said Fazl Barez, a elder researcher astatine the University of Oxford specialising successful AI information and governance.

However, helium cautioned that it was important to separate betwixt AI “automating tasks and replacing full jobs”.

“What are the types of jobs that could beryllium lost? Those involving regular cognitive enactment are peculiarly exposed, but the standard remains uncertain,” Barez told Al Jazeera.

“AI-driven automation threatens to erode authorities taxation bases, little surviving standards, and disempower citizens.”

AI needs to payment ‘everyone’

Despite concerns astir the AI revolution, Lamanauskas said helium did not judge determination was a “possibility conscionable to halt AI” oregon halt its accelerated advancement.

“AI is present to stay, and AI is present to alteration our lives,” helium said, noting that 104 countries present person AI strategies, compared with conscionable 4 a decennary ago.

“AI is not a exertion that tin beryllium controlled by 1 idiosyncratic oregon organisation oregon 1 company,” the UN authoritative added.

Nevertheless, helium called for greater concern successful integer infrastructure, acquisition and regularisation to guarantee processing countries tin enactment successful helping signifier the technology.

“The mode guardant is truly to put to marque definite that the AI benefits everyone, to recognise and admit the fears and anxieties that radical feel,” helium said.