ITU's Lamanauskas warns AI divide deepens as 2.2 billion offline

Lamanauskas's framing goes beyond the familiar digital divide. Because AI touches healthcare, education and industry, he argues that a gap in access becomes a gap in opportunity, and from there a deeper socioeconomic divide. The numbers he cites are not abstract: Africa and Latin America together hold 3.3 percent of global data center capacity, 2.2 billion people have no internet access at all, and only 7 percent of African languages are represented in AI translation tools.

The adoption gap is already visible. Roughly 29 percent of working-age people in developed countries use generative AI, against 16 percent in developing ones, which suggests the tools are spreading fastest where infrastructure and skills already exist. Lamanauskas puts the price of catching up at an estimated $2.8 trillion in investment by 2030, a figure that sets the scale of the policy challenge rather than a specific spending plan.

On jobs, he points to projections of 90 million displaced against 170 million created, with 40 percent of workplace skills shifting by the end of the decade. The net is positive in his account, but the churn is not evenly felt. His conclusion is not to slow AI down — he says AI is here to stay — but to invest so that benefits reach everyone and to acknowledge the fears people feel.