AI Compliance Training for Finance Teams: What Regulators Expect in 2026

AI Compliance Training for Finance Teams: What Regulators Expect in 2026

Excerpt: What the FCA, CBI, SEC, MAS and other regulators expect from financial services firms on AI governance, model risk, and staff training for AI tools.

Introduction: Regulatory Expectations Around AI Are Crystallising

For several years, regulators in major financial services jurisdictions issued general statements about the importance of responsible AI use. In 2026, that posture has sharpened considerably. The FCA, Central Bank of Ireland, SEC, MAS, and the EU (through the AI Act) have moved from principles to expectations — and financial services firms are now being asked to demonstrate that their AI governance frameworks are operational, not aspirational.

A critical component of any credible AI governance framework is staff training. Regulators across jurisdictions are clear that AI tools in financial services must be overseen by staff who understand those tools, their limitations, and the controls required to use them responsibly. This article sets out what each major regulator currently expects, what AI compliance training for finance teams should cover, and how Learnsignal can help your firm meet these expectations.

FCA: AI and Machine Learning Guidance

The FCA has been among the most active of global regulators in developing AI governance expectations for financial services firms. Its AI and machine learning guidance, first articulated in FS22/1 and developed through subsequent discussion papers and supervisory statements, sets out a framework for responsible AI use that has direct implications for staff training.

Key FCA Expectations

The FCA expects firms to be able to explain how AI systems make decisions — the explainability requirement. It expects appropriate human oversight of AI-generated outputs, particularly where those outputs are used in consumer-facing decisions. It expects firms to manage model risk — including the risk of bias, data quality issues, and model drift. And under Consumer Duty, it expects firms to be able to demonstrate that AI tools used in consumer-facing processes are producing outcomes that meet the standard of good consumer outcomes.

Implications for Staff Training

For FCA-regulated firms, AI compliance training should cover: what the firm's AI governance policy requires of staff who use AI tools; how to document AI-assisted decisions to meet explainability expectations; how to identify and escalate potential model risk issues; Consumer Duty implications for AI use in customer-facing processes; and SM&CR responsibilities for senior managers overseeing AI systems. Compliance teams need specific training on how AI intersects with the firm's regulatory obligations under the FCA rulebook.

Central Bank of Ireland: AI Governance in Regulated Firms

The Central Bank of Ireland (CBI) has been developing its supervisory approach to AI in financial services. The CBI has signalled increasing interest in how regulated firms are governing AI use, with particular focus on individual accountability frameworks and the competence of those overseeing AI systems.

IAF and AI Competence

Ireland's Individual Accountability Framework (IAF) — which extends Fitness and Probity obligations to a wider range of senior individuals — has implications for AI governance. Prescribed Responsibility holders and Senior Executive Function holders who oversee functions where AI is used are expected to have the competence to do so effectively. This does not require technical AI expertise, but it does require sufficient AI literacy to exercise meaningful oversight. CBI-supervised firms should ensure that relevant senior individuals have completed appropriate AI governance training.

MCC Requirements for Financial Advisers

Minimum Competency Code (MCC) requirements in Ireland apply to those providing financial advice. As AI tools are used more widely in financial advice processes, the CBI may develop specific expectations around competence in AI-assisted advisory processes. Finance teams in CBI-regulated firms should monitor CBI guidance in this area and ensure training programmes can adapt as expectations evolve.

SEC and FINRA: AI Expectations in the United States

In the United States, the SEC and FINRA have both issued guidance and enforcement actions related to AI use in financial services.

SEC Guidance on AI

The SEC has focused on AI-related disclosure obligations — particularly the requirement that registered investment advisers and public companies disclose material AI-related risks and governance practices. The SEC has also signalled concern about AI washing (overstating AI capabilities) and the use of predictive data analytics that may create conflicts of interest. For finance teams at SEC-regulated firms, training should cover AI-related disclosure obligations, conflicts of interest in AI use, and recordkeeping requirements for AI-assisted investment decisions.

FINRA's Approach

FINRA has issued guidance on AI use by broker-dealers, focusing on supervision — the requirement that AI systems used in customer communications, suitability analysis, or trading are subject to adequate supervisory controls. FINRA's AI guidance emphasises that firms remain responsible for AI-generated outputs under existing supervisory rules. Compliance training for FINRA-regulated firms should cover how AI tools fit within existing supervisory frameworks and what additional controls are required.

MAS9 Responsible AI in Singapore's Financial Sector

The Monetary Authority of Singapore (MAS) has published its FEAT (Fairness, Ethics, Accountability, and Transparency) principles for AI in financial services — among the most developed AI governance frameworks in any financial services jurisdiction. MAS has also participated in the development of the Veritas framework for responsible AI in financial services, which provides practical tools for assessing AI fairness and ethics.

MAS Expectations for Staff

MAS expects AI-supervised staff at regulated firms to be able to demonstrate understanding of the FEAT principles, model risk management practices, and how AI decisions are documented and explained. The MAS-IBF Skills Framework for Financial Services includes AI governance as a competency area. Finance teams at MAS-regulated firms benefit from AI compliance training aligned to the FEAT framework and MAS's expectations for model governance.

EU AI Act: Implications for Finance Teams

The EU AI Act, which came into full effect in phases from 2024, classifies AI systems used in financial services into risk categories with corresponding obligations. Many AI systems used in finance — credit scoring, fraud detection, insurance pricing, certain advisory systems — may be classified as high-risk under the Act, with significant governance requirements including staff training obligations.

Training Requirements Under the EU AI Act

For high-risk AI systems, the EU AI Act requires that those deploying or using the system ensure users have received adequate training on the system's intended purpose, its limitations, and the risks of misuse or over-reliance. This creates an explicit regulatory training obligation for finance teams using high-risk AI systems in EU-regulated entities. Finance teams in EU jurisdictions (including Irish IFSC firms, for example) need to map their AI systems against the Act's risk categories and ensure appropriate training is in place.

What AI Governance Training Should Cover for Compliance Staff

Compliance teams in financial services firms need AI training that goes beyond general AI literacy to cover specific governance and regulatory dimensions:

• AI governance frameworks: what a firm-level AI governance policy should contain and how to implement and monitor it
• Model risk management: the principles of model risk management (model validation, model approval, ongoing monitoring) and how they apply to AI models
• Regulatory mapping: how relevant regulatory obligations (FCA, CBI, SEC, MAS, EU AI Act) apply to the firm's specific AI use cases
• Explainability and documentation: how to document AI-assisted decisions to meet regulatory explainability requirements
• Bias and fairness: how to assess AI systems for bias and ensure AI use does not result in unfair treatment of customers or clients
• Incident response: how to respond to and report AI-related incidents, including potential regulatory notification obligations

Frequently Asked Questions

Is AI compliance training now a regulatory requirement for financial services firms?

In an increasing number of jurisdictions, yes — either explicitly (as under the EU AI Act's training obligations for high-risk AI systems) or implicitly (as a component of a credible AI governance framework that regulators expect firms to demonstrate). FCA-regulated firms are expected to evidence that staff using AI tools are appropriately trained as part of their AI governance arrangements. The direction of travel is towards increasingly explicit training requirements across all major jurisdictions.

How do we map our AI systems against regulatory risk categories?

Start by inventorying all AI systems and AI-enabled tools in use across your finance and compliance functions. For each system, assess: what decisions or outputs it generates, who those outputs affect, and what the consequence of an error or biased output would be. Map each system against the EU AI Act's risk classification framework and the relevant jurisdictional guidance (FCA, SEC, MAS, etc.). This mapping exercise typically requires input from both technology and compliance teams and should be reviewed annually or when new AI tools are deployed.

What should an AI governance training programme for compliance staff include?

A comprehensive AI governance programme for compliance staff should cover: AI fundamentals (what AI is and how it works at a non-technical level); the firm's AI governance policy and staff responsibilities under it; jurisdiction-specific regulatory expectations; model risk management principles; explainability, documentation, and audit trail requirements; bias assessment and fairness obligations; and incident response procedures. This is typically delivered as a full-day programme, with annual updates to reflect regulatory developments.

How should firms document AI compliance training for regulatory purposes?

Maintain training completion records for all staff who use AI tools, specifying the course completed, date, provider, and learning outcomes covered. For high-risk AI systems under the EU AI Act, maintain records sufficient to demonstrate that users received adequate training before using the system. In SM&CR jurisdictions, ensure that relevant senior managers' AI governance training is documented as part of their fitness and propriety records. Learnsignal provides completion certificates and training records suitable for regulatory documentation.

What is the risk of not providing AI compliance training?

The risks are both regulatory and operational. Regulatory risk: firms that cannot demonstrate adequate AI governance training may face supervisory scrutiny, remediation requirements, or enforcement action in the event of an AI-related incident. Operational risk: untrained staff are more likely to use AI tools in ways that create data privacy incidents, produce unverified AI outputs in regulated documents, or fail to identify model risk issues before they cause harm. AI compliance training is a risk management investment, not just a regulatory box-tick.

How frequently should AI compliance training be updated?

AI capabilities and regulatory guidance are both evolving rapidly. AI compliance training should be reviewed and updated at minimum annually, with interim updates when significant regulatory guidance is published (for example, major FCA papers, EU AI Act implementing regulations, or SEC enforcement actions with training implications). Staff should receive updated training when material changes are made to the firm's AI governance policy or when significant new AI tools are deployed.

Ensure Your Finance Team Meets Regulatory AI Training Expectations

Learnsignal designs AI compliance training programmes for financial services firms, aligned to FCA, CBI, SEC, MAS, and EU AI Act requirements. Our programmes cover AI governance, model risk, explainability, data privacy, and role-specific regulatory obligations — everything your finance and compliance teams need to meet regulatory expectations.

For bespoke AI compliance training for your finance team, visit our corporate training page to discuss your firm's regulatory context and training requirements.