Overview
This deep dive examines the latest developments in artificial intelligence, including emerging adoption patterns, key risk watch points, evolving policy and supervisory guidance, and research on the economic impact of AI.
State of play
Artificial intelligence adoption in financial services is accelerating at a rapid pace. Surveys by regulatory authorities in 2026 report AI use at the large majority of institutions, with adoption of up to 90% or more in several jurisdictions and the highest uptake among banks, insurers and larger firms. Generative AI accounts for most of the latest increase, and agentic systems that plan and execute multi-step workflows with limited human involvement are entering early deployment. Use remains concentrated in lower-risk applications, however. Most reported use cases are internal — drafting, summarisation, translation, coding support, internal search and adviser or call-centre support — while customer-facing, investment and core decisioning applications remain limited, more tightly controlled and subject to human validation. A large share of generative AI use cases are still at proof-of-concept stage, and agentic deployments are confined to bounded tasks with approval checkpoints for consequential actions. The constraints on scaling are consistent across surveys: business value that is difficult to evidence, data quality and legacy integration, dependence on a small number of external model and cloud providers, uncertainty over cross-border data use, and shortages of staff combining AI, risk and regulatory expertise.
Authorities’ attention has concentrated on the potential of AI to increase the speed, scale, interconnectedness and opacity of risk transmission across the…
View allView less
Artificial intelligence adoption in financial services is accelerating at a rapid pace. Surveys by regulatory authorities in 2026 report AI use at the large majority of institutions, with adoption of up to 90% or more in several jurisdictions and the highest uptake among banks, insurers and larger firms. Generative AI accounts for most of the latest increase, and agentic systems that plan and execute multi-step workflows with limited human involvement are entering early deployment. Use remains concentrated in lower-risk applications, however. Most reported use cases are internal — drafting, summarisation, translation, coding support, internal search and adviser or call-centre support — while customer-facing, investment and core decisioning applications remain limited, more tightly controlled and subject to human validation. A large share of generative AI use cases are still at proof-of-concept stage, and agentic deployments are confined to bounded tasks with approval checkpoints for consequential actions. The constraints on scaling are consistent across surveys: business value that is difficult to evidence, data quality and legacy integration, dependence on a small number of external model and cloud providers, uncertainty over cross-border data use, and shortages of staff combining AI, risk and regulatory expertise.
Authorities’ attention has concentrated on the potential of AI to increase the speed, scale, interconnectedness and opacity of risk transmission across the financial system. Four supervisory points dominate at present. The first is cyber risk. Frontier models can now discover and exploit software vulnerabilities autonomously, shortening the time between a weakness becoming known and its exploitation and lowering the cost and skill required for sophisticated attacks. Global bodies and national supervisors have raised their assessments of systemic cyber risk accordingly. The second is fraud. Generative AI makes scams more persuasive, personalised and scalable: deepfakes, synthetic identities and conversational bots can impersonate trusted people and institutions and sustain convincing interactions at volume. Exposure is growing on the consumer side as well, as more people take financial information from general-purpose AI tools that carry none of the protections of regulated advice. The third is concentration. A small number of model, cloud and software providers support critical functions across many institutions, creating shared points of failure and leaving firms with limited visibility into subcontractors, training data and hosting locations. Where the leading providers sit outside the home jurisdiction, authorities increasingly treat this as a strategic dependency. The fourth is financial stability. Common models could synchronise trading and redemption behaviour under stress, and the AI investment boom now features as a correction scenario in stress tests alongside credit and market shocks: capital expenditure is projected to exceed USD 4 trillion by 2029, funding is shifting towards bonds, private credit and securitisation, and circular investment and supply relationships among AI firms obscure underlying demand and provide channels for shocks to spread.
The policy and supervisory response is starting to converge. On regulatory architecture, most authorities apply existing governance, model risk, outsourcing, operational resilience and conduct frameworks to AI and clarify how they apply, rather than creating AI-specific regimes. The EU, where horizontal AI legislation applies alongside financial regulation, and several Asian jurisdictions with dedicated financial-sector AI guidelines are the main exceptions. On supervisory expectations, a common core has emerged across jurisdictions: visibility of AI use, risk tiering by materiality and autonomy, lifecycle controls, human accountability and third-party assurance, with more demanding requirements where AI affects customers, markets or critical operations. On generative and agentic AI, new risks are being translated into concrete controls, from records of prompts and configurations and limits on autonomous execution to runtime safeguards and AI-specific cyber testing. Agentic systems are the focus of the next wave of guidance, with several authorities planning consultations in 2027. On supervisory practice, authorities are institutionalising oversight through controlled testing environments, practical implementation toolkits, capacity building and public-private dialogue.
Evidence on the economic impact remains partial. Productivity gains are measurable for specific tasks and among adopting firms, but intensive use is rare and aggregate effects are not yet visible in the data. Labour-market effects are selective rather than systemic. Total employment has so far been little affected, while hiring of young and entry-level workers has slowed in the occupations most exposed to generative AI. The clearest macroeconomic effect to date is on investment. AI-related capital expenditure has increased substantially, driven by data centres, semiconductors and related infrastructure, and is increasingly financed through debt and capital markets rather than internal cash flow. Research points to a risk of overinvestment if productivity gains fall short of expectations. AI production remains concentrated in a small number of firms and economies, with most countries dependent on cross-border access to computing power, models and other inputs.
Recent developments
(Last update: )
Guidance has firmed up at global and national level in October. The Financial Stability Board (FSB) finalised its 12 sound practices for responsible AI adoption in mid-October, which set out proportionate, non-binding expectations for organisation-wide AI governance and for risk management across the AI lifecycle, and in its latest letter to G20 finance ministers and central bank governors Chair Andrew Bailey pointed to further work on AI-related risks, including the cyber risks posed by frontier models, in 2027. Across Asia, the Monetary Authority of Singapore issued its final Guidelines on AI Risk Management in early October, with effect from October 2027, and announced plans for a consultation focused on agentic AI systems in 2027. This was followed by the release of the Accountability, Resilience and Trust (ART) framework by the Hong Kong Monetary Authority, which combines updated supervisory guidance with a shared training-data platform and a planned governance framework for AI…
View allView less
Guidance has firmed up at global and national level in October. The Financial Stability Board (FSB) finalised its 12 sound practices for responsible AI adoption in mid-October, which set out proportionate, non-binding expectations for organisation-wide AI governance and for risk management across the AI lifecycle, and in its latest letter to G20 finance ministers and central bank governors Chair Andrew Bailey pointed to further work on AI-related risks, including the cyber risks posed by frontier models, in 2027. Across Asia, the Monetary Authority of Singapore issued its final Guidelines on AI Risk Management in early October, with effect from October 2027, and announced plans for a consultation focused on agentic AI systems in 2027. This was followed by the release of the Accountability, Resilience and Trust (ART) framework by the Hong Kong Monetary Authority, which combines updated supervisory guidance with a shared training-data platform and a planned governance framework for AI agents built around intent-anchored accountability, a registry of material AI agents and effective intervention mechanisms.
Supervisory follow-up on the cyber risks posed by frontier AI has shifted from an initial wave of warnings to more specific action points. Banks supervised by the European Central Bank (ECB) must submit their AI cyber-risk action plans by 31 October. Sweden’s Finansinspektionen has opened on-site reviews of how firms prevent, detect and manage AI-enabled attacks. Thailand’s Office of Insurance Commission ran two sector-wide cyber exercises with 67 insurers in late September, built around AI-related threats and failures at external providers.
The financing of the AI build-out has moved to the centre of financial stability assessments. The International Monetary Fund’s (IMF) Global Financial Stability Report, released against the backdrop of the Annual Meetings in Bangkok, names a correction in AI-related assets as one of three near-term triggers for market turmoil, and its global stress test finds that banks holding 22% of global banking assets would fall below a 7% CET1 ratio under a scenario combining an energy shock with an AI-led correction. An IMF technical note projects AI capital expenditure above USD 4 trillion by 2029, nearly three quarters of it by hyperscalers, and documents 14 circular financing deals in which about USD 300 billion of equity supported USD 1.4 trillion of future revenue commitments. A Bank for International Settlements bulletin finds that other AI firms account for 55% of investment into AI firms, almost half of it by value between firms that are also each other’s suppliers. Opening the annual conference of the European Systemic Risk Board, which she chairs, ECB President Christine Lagarde called for macroprudential monitoring of three AI threats: autonomous agents in trading, AI-enabled attacks spreading through shared technology, and geopolitical restrictions on access to frontier models.
Timeline of developments
European Central Bank blog says intensive AI use remains rare among euro area firms despite adoption exceeding 70%
An European Central Bank blog, based on SAFE survey data, said AI adoption among euro area firms exceeded 70% in late 2025 but only 7% used it intensively. The analysis linked intensive use to innovation-focused firms, peer pressure and broader financing needs, especially bank lending. It identified skill shortages, limited business usefulness and system incompatibility as the main barriers to deeper adoption.
Reserve Bank of India launches consultation on model risk management guidance covering AI and third party models
The Reserve Bank of India has issued draft guidance on model risk management for consultation. It sets lifecycle-wide expectations for governance, oversight and controls for all models used by a wide range of regulated entities, including third party and artificial intelligence or machine learning models. Comments are due by July 24, 2026.
Bank of Italy hosts G7 Cyber Expert Group meeting on cyberattack coordination and emerging technology risks
The Bank of Italy hosted a G7 Cyber Expert Group meeting in Palermo focused on cybersecurity and operational resilience in finance. Discussions covered lessons from a recent cross-border cyberattack coordination exercise and the risks associated with artificial intelligence and quantum computing. The meeting underscored the need for international coordination to manage operational interdependencies and emerging vulnerabilities.
Council of Financial Regulators flags elevated geopolitical and cyber risks while backing unchanged macroprudential settings
The Council of Financial Regulators said Australia's financial system remains resilient and supported APRA's decision to keep macroprudential settings unchanged. It warned that geopolitical tensions and frontier AI are increasing operational and cyber risks, and expects firms to strengthen contingency planning and cyber defences. The Council also reviewed progress on the 2026 IMF FSAP, cash distribution reform and cross-agency regulatory streamlining.
South Korea Financial Services Commission launches taskforce on capital market infrastructure and targets October settlement roadmap
The South Korea Financial Services Commission has launched a taskforce to modernize capital market infrastructure, combining trading and settlement reforms with wider adoption of AI and blockchain. Priorities outlined at the kickoff meeting include a roadmap for shortening the settlement cycle by October 2026, extended trading hours, AI-based market surveillance and removal of regulatory barriers to AI use. Regular meetings will review these measures and related regulatory hurdles.
Central Bank of Russia reports June business climate index fell to 1.1, first AI survey finds one in seven firms use the technology
The Central Bank of Russia’s June business survey showed the Business Climate Index fell to 1.1 points from 1.7 in May as companies became more cautious about the next three months. Price expectations declined for a fifth straight month. A first central bank survey on artificial intelligence found one in seven enterprises uses AI, led by service companies.
World Bank report says AI could raise Poland's real GDP by 1.3% to 12.1% by 2035
The World Bank said in a new report that AI could raise Poland's real GDP by 1.3% to 12.1% by 2035, with productivity gains starting within three years. It said the scale of the benefit will depend on investment, worker adaptation and supportive policies, with wider AI adoption, stronger managerial capacity and reskilling measures all critical to the outcome.
European Central Bank analysis finds AI has reduced US job growth in high substitution risk occupations without a clear wage effect
The European Central Bank published research indicating that AI has already shifted US job growth away from occupations with high substitution risk, even though the overall employment effect remains uncertain. It estimates that high-risk occupations grew around 15 percentage points less than low-risk occupations between 2019 and 2025, with no significant wage-growth gap yet visible. High-risk jobs fell by more than 4% over the period, while low-risk jobs rose by 13%.
Bank of England publishes research finding GPT-3.5 tracks short-horizon inflation attitudes but lacks a consistent inflation model
A Bank of England staff working paper finds GPT-3.5 can mimic UK household inflation perceptions at an aggregate level when given price signals, especially at short horizons. The paper also finds weak individual-level alignment, unreliable longer-term expectations and inconsistent sensitivity across inflation components. The framework is presented as a way to test LLMs for research and survey design.
European Central Bank outlines bank supervision simplification agenda and plans cyber resilience letter to CEOs
In remarks published by the European Central Bank, Frank Elderson said ECB Banking Supervision is pursuing simplification that cuts unnecessary complexity without weakening resilience, while arguing that fragmented EU banking markets rather than capital rules are the bigger drag on competitiveness. He highlighted faster approval processes, a planned dear CEO letter on cyber resilience and continued work on a digital euro pilot in 2027 with potential issuance readiness in 2029 if legislation is passed in 2026.
Italian Institute for Insurance Supervision publishes macroprudential committee assessment that global uncertainty remains the main risk to Italy’s financial system
The Italian Institute for Insurance Supervision published a macroprudential committee update saying the main risks to Italy’s financial system still come from a highly uncertain international environment, despite broadly solid financial conditions and orderly markets. The committee also highlighted operational and cyber risks linked to artificial intelligence, continued monitoring household holdings of certificates, and reviewed work on cryptoasset investments and benchmarks.
European Banking Authority reports EU EEA banks remain resilient while highlighting rising geopolitical cyber and non-bank risks
The European Banking Authority said EU and EEA banks remain resilient, supported by solid capital, liquidity, asset quality and profitability, despite higher uncertainty in financial markets. Its latest risk publications also warn of rising geopolitical, operational and cyber risks, along with growing interconnectedness between banks, non-bank financial institutions and private credit markets.
China's National Financial Regulatory Administration issues 32 point guidance on safe AI development and use in banking and insurance
China's National Financial Regulatory Administration has issued guidance for banks and insurers on the safe development and use of AI. The 32 measures require full life cycle governance, stronger data and computing arrangements, and AI risk controls including classification, high risk application entry management and human oversight. Supervisors will oversee implementation, monitor compliance risks and take action against violations.
China Banking and Insurance Regulatory Commission issues guidance on safe AI development and use in banking and insurance
The China Banking and Insurance Regulatory Commission has issued guidance requiring banks and insurers to govern AI through board-level oversight, full life cycle controls and risk-based classification of applications. It sets expectations for model, data, computing, transparency and cybersecurity controls, including labelling of AI-generated content and human intervention for high-risk uses. Firms using generative AI in public-facing or high-risk scenarios must report this to the regulator.
Dubai International Financial Centre launches consultation on data protection amendments for AI systems
The Dubai International Financial Centre has opened a consultation on amendments to its Data Protection Regulations focused on AI-enabled and data-driven systems. The proposals would strengthen safety and privacy-by-design requirements, clarify certification obligations and the Autonomous Systems Officer role, and let the Commissioner recognise accreditation and certification schemes. Comments are due by 18 July 2026.
Financial Stability Board convenes roundtable on audit quality and ownership changes in the global audit industry
The Financial Stability Board held a roundtable on the financial stability implications of structural change in the global audit industry. Discussions focused on audit firm ownership changes linked to technology investment, including artificial intelligence, and on how those shifts may affect audit quality and the adequacy of existing standards. Participants also discussed ways to address a recent adverse trend in audit inspection findings.
International Organization of Securities Commissions publishes SupTech report showing wider supervisory use but persistent cyber and funding constraints
The International Organization of Securities Commissions published a SupTech report based on a survey of 49 jurisdictions, finding that supervisory technology is becoming embedded in core oversight functions. Use is strongest in investor protection and capital markets supervision, with digital assets emerging as a growth area. Cyber risk, funding limits and weak long-term workforce planning remain the main constraints.
South Korea Financial Services Commission updates financial sector AI guidelines and outlines phased sandbox testing
The South Korea Financial Services Commission has issued revised AI guidelines for the financial sector, setting out seven principles for AI use by financial companies and fintech firms. The self-regulatory framework takes effect on June 22, 2026, with supporting risk management and security guidance to follow. The commission also outlined further work on regulatory easing, AI accountability rules and phased sandbox testing for AI agents in the second half of 2026.
China Securities Regulatory Commission sets out market reforms, wider STAR Market access for AI and a RMB foreign exchange futures pilot
In remarks at the Lujiazui Forum, the China Securities Regulatory Commission outlined a new reform agenda for listings, products and cross-border access. It plans broader STAR Market access for artificial intelligence and other hard-technology issuers, faster M&A and refinancing tools including shelf issuance, active ETFs and commercial real estate REITs, and work toward a RMB foreign exchange futures pilot. The commission also signaled tighter oversight of AI-related market abuse and unlawful cross-border business.
Dubai International Financial Centre launches banking report warning lagging transformation could cut industry profit pools by USD 170bn by 2030
Dubai International Financial Centre has released a banking report saying resilience and AI-led operating model change will determine banks' long-term performance more than scale or legacy. It warns that without decisive transformation, industry profit pools could fall by USD 170bn by 2030, as challenger banks force incumbents to match faster, more personalized and lower-cost models. The report is the second in DIFC's four-part 2026 Future of Finance series.
Central Bank of Russia publishes first AI information security recommendations for financial institutions
The Central Bank of Russia has issued its first recommendations on information security for financial institutions using AI. The guidance maps AI-related risks and cyberattack tactics and recommends human confirmation where AI is used in high-risk critical processes such as payment transactions. It also calls for internal AI threat models and security policies and highlights vendor security practices, including bug bounty participation.
European Central Bank publishes working paper finding firms react to domestic but not foreign peer AI investment
The European Central Bank has published a working paper, representing the authors’ views, finding that firms underestimate peers’ AI investment and increase their own planned AI spending when given accurate information. The effect is driven by domestic peers, with a 1 percentage point rise in expected domestic peer adoption increasing a firm’s own expected AI investment rate by 0.570 percentage points, while foreign peer effects were not statistically significant. Treated firms planned to allocate 10.13% of investment to AI versus 8.33% in the control group.
Hong Kong Mandatory Provident Fund Schemes Authority launches AI chatbot for personalized MPF investment education
The Hong Kong Mandatory Provident Fund Schemes Authority has launched an AI chatbot on its MPF investment education website to give scheme members real-time, personalized information based on official MPFA resources. The tool explains MPF investment concepts, provides source references and visual aids, and suggests related questions. It does not provide investment advice, recommend funds, or assess suitability, and will later be extended to other education tools including the retirement planning mobile app.
Belgium Financial Services and Markets Authority warns frontier AI heightens cyber risk and tells DORA firms to strengthen controls
The Belgium Financial Services and Markets Authority has warned that frontier AI systems are making cyberattacks easier, faster and more scalable, requiring regulated firms to reassess ICT risk even if they were previously seen as lower-risk targets. It says DORA provides the core response framework and expects in-scope firms to strengthen asset inventories, patching, detection and incident response, and oversight of ICT service providers. The guidance is also recommended for firms outside DORA's scope.
Canadian Securities Administrators warn of rising fraud risks for older investors and urge use of trusted contacts
The Canadian Securities Administrators warned that older investors are increasingly exposed to fraud and financial abuse as digital platforms and AI make scams look more legitimate. It urged investors to verify registration, be wary of unsolicited and crypto-related offers, and avoid relying on apparent endorsements from public figures. The CSA also reminded investors to name a Trusted Contact Person on investment accounts as a safeguard against suspected abuse or exploitation.