Quick Read Summary
- At the IMF and World Bank meetings in Bangkok, IMF chief Kristalina Georgieva argued that AI’s economic gains depend on safeguards, infrastructure and a credible ethical framework.
- The warning connects AI governance to jobs, infrastructure, inequality and financial stability rather than treating safety as a purely technical issue.
At the IMF and World Bank meetings in Bangkok, IMF chief Kristalina Georgieva argued that AI’s economic gains depend on safeguards, infrastructure and a credible ethical framework.
Georgieva said AI’s risks could outweigh its opportunities without appropriate safeguards and oversight.
She argued that access to digital infrastructure and electricity is necessary for AI’s benefits to be broadly shared.
AI’s economic effect will vary by occupation, region and the speed at which employers reorganise work. Some tasks may become faster or less expensive, while other roles could see demand fall or shift toward supervision, data quality and customer-facing judgment. If productivity benefits are concentrated in some firms, the gains may not automatically translate into broadly shared wages or stronger job security. Policymakers need reliable data on adoption, wages, job transitions and business investment rather than assuming either universal displacement or universal growth.
Other participants raised concerns about job displacement, financial imbalances and the scale of investment required to provide expected productivity gains.
The distribution of AI benefits depends on access to reliable power, connectivity, education and capital, not only the quality of a model. Countries and businesses that lack those inputs may see costs rise before productivity gains appear. Oversight also needs to account for privacy, security and accountability, because failures can reduce public trust and slow adoption. The policy challenge is to reduce foreseeable harm while allowing useful applications to reach sectors beyond the largest technology companies.
The IMF’s concern is not that AI has no economic value, but that gains may be distributed unevenly and may not arrive as quickly as investment expectations imply. Firms with computing resources, data and skilled workers can adopt AI faster than smaller businesses or regions with unreliable electricity and limited connectivity. If productivity gains are concentrated while jobs are displaced or capital spending outruns revenue, the transition can create social and financial pressure. Safeguards matter because trust, security and clear accountability influence whether organisations can deploy AI in sensitive sectors. Policy choices will shape how much of the potential benefit reaches workers and consumers rather than only technology owners.
Safeguards are not limited to technical model testing. They can include worker training, employment support, competition policy, clear rules for sensitive uses and systems that let people challenge automated decisions. Financial authorities may also need to consider concentration risks if many firms rely on a small set of AI providers or if automated systems create correlated behaviour during market stress. The IMF chief’s warning should be read as a call to manage uncertainty, not a definitive forecast that the harms will exceed the benefits. Outcomes will depend on how the technology is deployed and how institutions respond.
The policy challenge is to allow useful adoption while making the transition less damaging for workers and consumers. Credible measurement, transparent accountability and the ability to appeal consequential decisions are central to that balance.
Watch how governments translate broad calls for safeguards into concrete standards, oversight and support for workers and smaller economies.