A survey of 1,267 Indian employers and 2,541 employees reveals a sharp divergence: 66% of companies pay high premiums for AI-skilled roles, yet 54% of workers in AI-exposed jobs saw their pay stagnate or fall over the past year. AI-related hiring rose 16% year-on-year even as overall technology recruitment contracted 3%. This bifurcation is not confined to India; it reflects a global pattern in which the promise of artificial intelligence is reshaping labour markets unevenly.

The dynamic is driven by a push to deploy generative AI without expanding headcount. Start-ups from Silicon Valley to Paris are using AI agents to handle coding, customer service and administrative tasks, allowing them to grow revenue while staying lean. A study by payments platform Stripe notes an explosion of single-person companies since late 2024, as AI fills needs that once required hires. Larger firms, too, are reallocating budgets: they bid up salaries for external AI specialists but tell existing employees to expect muted appraisals, widening the gap between new recruits and tenured staff.

In Latin America, economists estimate AI could add 1.1–1.7 trillion dollars in value and lift annual productivity by up to 2.3%, yet only 6% of organisations report capturing significant benefits. Nobel laureate Christopher Pissarides, speaking in Rio de Janeiro, argued that fears of mass unemployment are overblown—AI is more an assistance tool than a replacement—but warned of regional inequality: 60% of AI investment concentrates in elite metropolitan corridors. In Sweden, a survey of academics in Östergötland shows optimism about AI’s effect on work has slipped to 65%, with the young markedly more pessimistic. Mexican industry, meanwhile, is betting on digital infrastructure to integrate manufacturing, logistics and services, as campaigns like “Lo Hecho en México Siempre Gana” promote the country’s capacity to absorb nearshoring through technology.

The immediate test is whether companies will invest in reskilling their existing workforces, as Indian HR analysts urge, or continue to source AI talent externally at the expense of loyalty and institutional knowledge. In Sweden, the issue will land on the desk of the government that emerges from autumn elections, with unions and professional bodies demanding lower thresholds to first jobs and lifelong learning frameworks. Without such interventions, the current dichotomy risks hardening into a structural feature of the AI-era labour market.