Most central banks are using AI, but only 4% reported high compliance-monitoring capacity
An early-2026 BIS/IFC survey of 65 central-bank jurisdictions and institutions found AI projects moving into production while governance, data readiness, specialist staffing and monitoring still lag. The results are self-reported and do not measure productivity or safety outcomes.
The short version
- What happened
- The early-2026 survey covered 65 responding jurisdictions and institutions. More than two-thirds reported AI projects in production, including small-scale use, compared with about half in 2024.
- Who it affects
- Central-bank AI can affect the staff producing statistics, supervising firms and supporting policy analysis. Clear inventories, accountable human review and tested controls matter before model outputs influence decisions that reach households or businesses.
- What remains unproven
- Self-reported institutional survey rather than an independent audit
Keep in mind
Self-reported institutional survey rather than an independent audit
65 responding jurisdictions and institutions; question denominators vary
Respondent composition differs somewhat from the 2024 survey
No system-level accuracy, bias, security or incident measurements
No productivity, policy-quality or public-service outcome evaluation
Authors' views do not necessarily represent the IFC, its members or the BIS
Source analysis · 1 linked source. Primary analysed source date: 28 October 2026.
Keep following the evidence
Join The Impact Brief founding list for selected AI reports, their effect on people and the evidence limits. Read a sample first.
Read the full report · 3 min The evidence, methods, implications and limitations
AI use is spreading faster than some readiness controls
The Irving Fisher Committee surveyed 65 central-bank jurisdictions and institutions in early 2026. More than two-thirds said they had AI projects in production, including small-scale deployments, up from about half in 2024. Almost two-thirds now treat AI as a priority. Organisational strategies were already in place at 38% of respondents and being drafted at another 31%.
The control picture was less mature. Only 4% reported high capacity to monitor compliance with AI guidance; 54% described moderate capacity, 27% said monitoring was under development and 15% called it limited. Data readiness was also uneven: 21% had a strategy in place, 41% planned one and 29% did not plan one. These gaps matter when central-bank tools touch statistics, supervision, market analysis or internal decisions.[1]
The survey maps readiness, not whether AI improved decisions
Infrastructure and people remain concentrated constraints. Sixty-five per cent reported dedicated AI infrastructure, and 73% employed AI professionals; among those, 81% said specialists represented less than 2.5% of staff. Most respondents allocated less than 5% of their budget to AI, while almost 80% cited IT costs as a concern. External models and providers therefore remain part of the operating picture.
These are self-reported institutional answers, not audited inventories. Response bases differ by question, and the 2024 and 2026 respondent groups are not identical, although the report says composition changes did not materially alter aggregate comparisons. The study does not show that AI made policy, supervision or services better—or that guidance was followed in practice. Independent system inventories, incident reporting and outcome evaluations would change the assessment.[1]
The distinction that matters
Readiness signals from the 65-response survey
Production use, strategy and monitoring are related but not interchangeable measures.
Use
AI projects in production, including small-scale deployment
More than two-thirds of respondents
Strategy
Organisation-wide AI strategy in place
38% of respondents
Monitoring
High capacity to monitor compliance
4% of respondents
Watch the background, then check the evidence
Illustrative photo · not a video frame
UNESCO · 11 min 27 sec · Background explainer
UNESCO Talks: Interview with 2024 Nobel laureate Professor Daron Acemoglu
YouTube loads only when you choose; playback starts when you press play.
Economist Daron Acemoglu discusses how choices about AI affect workers, inequality and society. Pair the interview with our comparison of task exposure, hiring and skills evidence.
Evidence check: An expert interview from 2024, offering a perspective on policy and work rather than a measured forecast of job losses.
Published 2024-10-14. Source checked 14 October 2026. Check captions on the publisher’s video page.
Read the AI and jobs guidePhoto: Dylan Gillis / Unsplash · Licence
Continue the story
Read next
Continue exploring ai governance: compare the finding and its evidence limits.
What does frontier AI change for banks?
Canada's federal prudential regulator says frontier AI is compressing cyber-defence time and increasing dependence on a small number of model and cloud providers. The 8 October update is a supervisory risk assessment, not evidence of a new bank failure, breach or enforcement action.
Read the next reportExplore the full source and evidence record
What this means for people
- Central-bank AI can affect the staff producing statistics, supervising firms and supporting policy analysis. Clear inventories, accountable human review and tested controls matter before model outputs influence decisions that reach households or businesses.
Global context
The respondent list spans advanced and emerging economies across regions, but institutions differ greatly in mandate, staffing, procurement and regulation. Aggregate percentages cannot tell a particular country whether its central bank has safe, effective systems or sufficient public accountability.
What the evidence does not yet show
- Self-reported institutional survey rather than an independent audit
- 65 responding jurisdictions and institutions; question denominators vary
- Respondent composition differs somewhat from the 2024 survey
- No system-level accuracy, bias, security or incident measurements
- No productivity, policy-quality or public-service outcome evaluation
- Authors' views do not necessarily represent the IFC, its members or the BIS
What to watch next
- Published inventories of central-bank AI systems and their purposes
- Independent audits and incident disclosures
- Evidence that compliance guidance changes practice
- Measured effects on staff workload and decision quality
- Procurement concentration and external-provider dependencies
Living evidence record
Impact record IAI-1JHL9YC
Evidence stage
Observed
Confidence
Supported
Reporting basis
Source analysis
Independent or research support
Not yet
Record status
Monitoring
Last checked
28 October 2026
Source trail
1 direct source across 1 source type.
People impact
Documented in this record.
Uncertainty
Limits and next checks are explicit.
Stages describe the evidence available—not whether a technology is good or bad. See the public method.
Single-source reporting disclosure
This record analyses one direct source. It can establish what Bank for International Settlements / Irving Fisher Committee published or reported, but it is not independent corroboration of every performance claim or predicted outcome. The confidence label will change only when broader evidence is added.
Evidence trail
Sources used for this report
Links checked 28 October 2026
Research question, topics and regions
How central banks describe their organisational, data, infrastructure, staffing and governance readiness for AI
Topics: Central banking · AI governance · Financial institutions · Compliance · Data readiness · Public-sector technology
Regions: Global
This report is labelled source analysis. We summarise and analyse source material in our own words; company statements remain attributed claims until independently supported. Translated summaries preserve the meaning of the original source and link back to it. Read our editorial standards.
Reader commentary
What does this mean where you are?
Tell us what you have seen at work, in your community or in everyday life. A question is welcome too. No account is needed; we review notes before publishing them.
Explore commentary across the portal →Published reader notes
0No published reader notes yet. You can start the evidence-led discussion above.
Prefer a private correction or response? Contact the newsroom.