AI Debt Collection in the UK: FCA Compliance Guide (2026)

The UK Debt Collection Market

The UK debt collection market is mature, heavily regulated, and increasingly open to AI adoption. With over 8 million UK adults experiencing problem debt and a consumer credit market exceeding £200 billion, the demand for effective and compliant collection is substantial.

UK debt collection operates differently from the US in several important ways. There is no equivalent of the FDCPA as a single comprehensive debt collection statute. Instead, the regulatory framework combines the Consumer Credit Act 1974 (as amended), FCA rules and guidance, the Consumer Rights Act 2015, the Equality Act 2010, UK GDPR and the Data Protection Act 2018, and common law principles. This creates a multi-layered compliance environment that AI must navigate.

The FCA authorizes and supervises all consumer credit activities in the UK, including debt collection. Firms must hold a consumer credit authorization from the FCA to engage in debt collection activities. Operating without authorization is a criminal offense. This licensing regime is more stringent than most US state licensing requirements and provides a baseline standard that all market participants must meet.

The UK market also has a strong tradition of free debt advice services - StepChange, Citizens Advice, National Debtline, and others provide free help to people in financial difficulty. AI systems must be configured to reference these services and pause collection when a debtor is engaging with a debt advice provider.

FCA Regulatory Framework for Debt Collection

The FCA's approach to debt collection regulation is outcomes-based rather than rules-based. Where the US FDCPA provides specific prohibited practices (no calling before 8 AM, no threatening arrest), the FCA requires firms to achieve good outcomes for customers. This is simultaneously more flexible and more demanding - there are fewer bright-line rules to follow, but the standard for acceptable conduct is higher.

The CONC (Consumer Credit) sourcebook contains the specific rules for debt collection conduct. CONC 7.3 is particularly important - it requires firms to treat customers in default or arrears with forbearance and due consideration. This means AI cannot simply demand payment. It must assess the customer's circumstances, offer appropriate options, and suspend enforcement activity when the customer is in genuine financial difficulty.

Consumer Duty and AI: The 2023 Rules

The FCA's Consumer Duty, which came into force in July 2023, represents the most significant shift in UK financial regulation in decades. It requires firms to deliver good outcomes for retail customers across four areas: products and services, price and value, consumer understanding, and consumer support.

For AI debt collection, the Consumer Duty creates specific obligations. The products and services outcome means that the collection process itself must be designed to produce good outcomes for customers - not just good outcomes for the collecting firm. This requires AI to offer genuine resolution options, not just push for maximum payment.

The consumer understanding outcome means AI must communicate in a way that customers can actually understand. No complex financial jargon, no fast-spoken legal disclaimers, no ambiguous statements about consequences. The AI must check for understanding and offer to explain further when the customer seems confused.

The consumer support outcome means customers must be able to get help when they need it. For AI, this means seamless escalation to human agents when the situation requires it, and availability through channels that work for the customer. If a customer says they cannot hear well on the phone and needs written communication, AI must accommodate that preference.

The Consumer Duty also requires firms to monitor outcomes and demonstrate that they are delivering good results for customers. AI analytics provide the data to prove compliance - average call duration, payment plan sustainability rates, customer satisfaction scores, and complaint volumes can all be tracked and reported to the FCA.

Treating Customers Fairly (TCF) Principles with AI

Treating Customers Fairly has been a cornerstone of FCA regulation since 2006. For debt collection, TCF translates into specific behavioral expectations that AI must embody.

Vulnerable Customer Identification and Handling

The FCA defines a vulnerable customer as someone who, due to their personal circumstances, is especially susceptible to harm. The FCA identifies four key drivers of vulnerability: health conditions (physical disability, mental health issues, severe illness), life events (bereavement, job loss, relationship breakdown), resilience (low financial resilience, over-indebtedness), and capability (low literacy, English as second language, digital exclusion).

AI must be able to identify vulnerability indicators during calls and adjust its approach accordingly. This is not a binary classification - vulnerability exists on a spectrum, and different vulnerabilities require different responses.

A customer who mentions they have recently lost their job needs immediate referral to income maximization support and a realistic assessment of their ability to pay. A customer who seems confused or has difficulty following the conversation may have cognitive issues that require simpler language and more patience. A customer who becomes distressed or mentions suicidal thoughts needs immediate escalation to a trained human handler with access to crisis support resources.

The FCA expects firms to have systems that identify, record, and respond to customer vulnerability. AI's natural language processing capabilities make it better at detecting certain vulnerability indicators than human agents - it can identify speech patterns associated with cognitive difficulty, emotional distress markers in tone and word choice, and contextual clues that suggest life events. However, the response to identified vulnerability must be genuinely helpful, not just a tick-box exercise.

Consumer Credit Act Requirements

The Consumer Credit Act 1974 (CCA) provides the statutory foundation for consumer credit regulation in the UK. For debt collection, several CCA provisions are directly relevant.

Section 77-79 gives debtors the right to request a copy of their credit agreement and statement of account. When a debtor makes this request, the creditor must comply within 12 working days. If the creditor cannot produce the agreement (common with older debts), collection activity must stop until it is produced. AI must recognize these requests, pause collection, and route the account for agreement production.

Section 87 requires a default notice before a creditor can take certain enforcement actions. The default notice must give the debtor at least 14 days to remedy the default. AI must verify that a valid default notice has been served before referencing enforcement actions in collection calls.

Section 140A-C allows courts to find credit relationships unfair to the debtor. This provision has been used successfully to challenge aggressive collection practices, disproportionate fees, and inadequate consideration of the debtor's circumstances. AI's consistent, documented, and fair approach to collection actually reduces the risk of an unfair relationship finding.

AI-Specific Compliance Considerations

The FCA has issued guidance on the use of AI and machine learning in financial services. While there is not yet a specific AI regulation for debt collection, the FCA's expectations are clear and are being applied through supervisory engagement.

The disclosure requirement is straightforward - AI must inform the customer at the beginning of the call that they are speaking with an AI system and offer the option to speak with a human agent. Unlike some jurisdictions where this is ambiguous, the FCA's transparency expectations make disclosure effectively mandatory.

Bias testing is particularly important for debt collection AI. If the AI system produces different outcomes for customers of different ethnic backgrounds, ages, genders, or other protected characteristics - even unintentionally - the firm faces both FCA enforcement risk and Equality Act liability. Regular fairness audits comparing AI outcomes across demographic groups are essential.

Implementation Guide for UK Debt Collectors

ICO and UK GDPR Data Protection

UK GDPR (the UK's post-Brexit version of the EU General Data Protection Regulation) applies fully to AI debt collection. The Information Commissioner's Office (ICO) is the relevant supervisory authority, and it has shown particular interest in AI-related data processing.

The lawful basis for processing debtor data with AI is typically legitimate interests (Article 6(1)(f)) for the collection activity itself, and contract performance (Article 6(1)(b)) where the debt arises from a contractual relationship. However, the legitimate interests assessment must balance the firm's collection interest against the debtor's rights - and the ICO has indicated that AI processing may require stronger justification than manual processing because of its scale and the reduced human oversight.

Automated decision-making provisions under Article 22 of UK GDPR are directly relevant. If AI makes decisions that have legal or similarly significant effects on individuals (such as determining payment plan terms, deciding whether to escalate to legal action, or assessing creditworthiness), the debtor has the right to not be subject to purely automated decisions and to request human intervention. AI systems must implement a meaningful human review mechanism for significant decisions.

For EU-based operations alongside UK collection, our guide on GDPR-compliant AI debt collection in Europe covers the additional requirements. And for broader context on how AI debt collection works internationally, see our comprehensive guide.