Designing KYC interview questions for smoother client onboarding

Two people sitting across a desk in a meeting room, reviewing documents together
TL;DR

KYC interview questions are the structured questions owner-managed services firms ask at client onboarding to verify identity, confirm beneficial ownership, and understand the nature of the engagement. The Money Laundering Regulations 2017 require customer due diligence from accountants, tax advisers, estate agents, and company service providers before work begins. The FCA and JMLSG expect a proportionate, risk-based approach: lighter questions for standard-risk clients, enhanced diligence only where the risk profile warrants it.

Key takeaways

- Firms in sectors covered by the Money Laundering Regulations 2017, including accountants, tax advisers, estate agents, and company service providers, have a legal obligation to conduct customer due diligence before client work begins. - The FCA and HM Treasury both expect a risk-based, proportionate approach: lighter questioning for standard-risk clients, enhanced due diligence only where the risk profile warrants it. - Copying full bank-grade KYC questionnaires creates unnecessary friction and conflicts with the MLR 2017's proportionality principle, which permits simplified processes for lower-risk clients. - AI tools can reduce onboarding friction through dynamic question routing, Companies House auto-population, and background sanctions screening, but firms remain fully responsible for outcomes and must build in human oversight. - All customer due diligence records, including AI-generated transcripts and structured onboarding data, must be retained for five years under the MLR 2017.

A client inquiry comes in. You send over your onboarding form. Two days later, you haven’t heard back. They’ve found someone who asked fewer questions. The form wasn’t wrong exactly, it just wasn’t calibrated for the relationship you were trying to start.

For owner-managed services firms that fall under the UK’s anti-money laundering rules, client due diligence isn’t optional. But the way you structure the questions is entirely within your control, and it matters more than many firms realise.

What are KYC interview questions?

KYC stands for Know Your Customer. The questions you ask at client onboarding to verify identity, confirm beneficial ownership, and understand the nature of the engagement. For regulated professional services firms, specifically accountants, tax advisers, estate agents, and company service providers, these questions are a legal requirement under the Money Laundering Regulations 2017. Done well, they collect what the law needs without making the client feel interrogated.

The Regulations require four core elements: identity verification, beneficial ownership confirmation, understanding the purpose of the business relationship, and ongoing monitoring. Enhanced due diligence applies for higher-risk clients, such as politically exposed persons or those connected to high-risk jurisdictions. The practical challenge is sequencing those requirements in a way that feels like a professional intake process, not a customs interview.

For a corporate client, the base set covers legal name, company number, registered address, nature of business, and confirmation of who ultimately owns or controls the entity. For an individual, it is full legal name, date of birth, residential address, and preferred contact details. Companies House integration can auto-populate the corporate fields, reducing what the client types directly.

Why does question design matter as much as compliance?

How you sequence and frame these questions determines whether clients reach the end of your onboarding process. Signicat’s 2022 survey found that 68% of UK and European consumers had abandoned at least one financial services application in the previous 12 months, with verification complexity among the key reasons. Owner-managed services firms aren’t banks, but the same friction pattern applies when questionnaires are over-long, unclear, or ask for information clients can’t readily provide.

The FCA’s Financial Crime Guide is explicit that firms should tailor customer due diligence to the client’s risk profile, not run every client through the same maximum-level process. HM Treasury’s National Risk Assessment makes the same point: proportionate, risk-based questioning, not a box-ticking exercise. That regulatory expectation supports the design decision you want to make anyway, asking more from higher-risk clients and less from straightforward ones.

The Standard Chartered fine of £102.2 million in 2019, for failures in customer due diligence across its correspondent banking relationships, illustrates the direction of regulatory risk. For an owner-managed firm, the lesson is to ask the right questions in the right order and verify answers with independent sources, rather than using questionnaire volume as compliance insurance.

Where will you actually meet KYC requirements as a services firm?

The Money Laundering Regulations 2017 cover a defined set of sectors. Estate agents, accountants, tax advisers, insolvency practitioners, auditors, and company service providers all fall within scope. If your firm sits in one of those sectors, you already have a legal obligation to conduct customer due diligence before work begins. The question is how you design that process, not whether it applies.

Firms outside those defined sectors still encounter KYC-style processes from the other direction. If your clients are in financial services, regulated industries, or large corporate supply chains, they will ask you to complete due diligence questionnaires as a condition of the business relationship. Understanding what a well-designed KYC process looks like helps you complete those forms accurately and identify when something unusual is being requested.

The JMLSG guidance for the UK financial sector sets out the risk-based approach in practical detail, covering when simplified due diligence is acceptable, when standard applies, and when enhanced is required. It is the clearest published benchmark for proportionate customer due diligence available to UK firms, and it is publicly accessible. Regulated or not, reading the sectoral guidance for your closest parallel sector is a useful starting point for calibrating your own onboarding questions.

When should you go deeper, and when can you keep onboarding light?

Standard due diligence covers the majority of client onboarding. For lower-risk clients, a straightforward process covering identity, beneficial ownership, and the nature of the engagement is sufficient under the MLR 2017. Enhanced due diligence is required in specific circumstances: politically exposed persons, clients with connections to high-risk jurisdictions, and situations where the transaction pattern or value is disproportionate. Knowing which category applies at intake is the critical design decision.

A practical trigger system keeps this manageable. At intake, you collect jurisdiction, the nature of the work, and a broad value band. Those inputs sort clients into standard or higher-risk categories before any human reviews the file. Higher-risk clients then receive an additional set of questions covering source of funds, source of wealth for individuals, and a more detailed description of the business relationship. Clients who fall outside those parameters go through a lighter process.

On the question of PEP status, the cleaner onboarding approach is to frame the question around specific role types rather than asking clients to self-identify as politically exposed. Asking whether any individual holds or has recently held a prominent public function draws on the FCA’s PEP definition and generates a more actionable response. Screen names against a maintained list in the background, and reserve the detailed follow-up questions for situations where that screening returns a flag.

How does AI fit into KYC question design?

AI adds the most value in KYC onboarding at the question-routing and verification stages. Dynamic forms surface only the follow-up questions that are logically necessary based on earlier answers, whether the client is an individual or a corporate entity, UK-based or connected to an overseas jurisdiction. That reduces the total number of questions a client sees without reducing the information you collect.

Document verification tools, such as those provided by Onfido, use AI to confirm identity documents are genuine and match the person presenting them, completing in minutes what manual review takes days to achieve. Sanctions and PEP screening through services like ComplyAdvantage or Refinitiv World-Check run in the background against client names, flagging potential matches for human review.

The FCA has been clear that firms using AI in financial crime controls remain fully responsible for outcomes and must ensure appropriate oversight, data quality, and the ability to explain decisions. The ICO’s guidance under UK GDPR Article 22 adds a further constraint: solely automated decisions with significant effects on individuals require explicit consent or a clear legal basis, plus a mechanism for human review. Build human oversight into the design before you start, rather than treating it as a compliance add-on.

The NCSC recommends logging and monitoring AI system outputs to detect anomalies and security issues. For onboarding tools specifically, that recommendation aligns with the MLR 2017’s five-year record retention requirement. If your tool generates a chatbot transcript or a structured data record, your existing customer due diligence retention obligations extend to those outputs.

Sources

- UK Government (2017). Money Laundering, Terrorist Financing and Transfer of Funds (Information on the Payer) Regulations 2017. The primary UK legal framework requiring customer due diligence for regulated professional services firms. https://www.legislation.gov.uk/uksi/2017/692/contents - FCA (2015, updated). Financial Crime Guide: A firm's guide to countering financial crime (FG15/6). Sets out the FCA's expectation for risk-based, proportionate customer due diligence, including understanding the nature and purpose of the business relationship. https://www.fca.org.uk/publication/finalised-guidance/fg15-06.pdf - HM Treasury (2020). UK National Risk Assessment of Money Laundering and Terrorist Financing. Establishes the proportionality principle for risk-based questioning and sectoral guidance across regulated firms. https://assets.publishing.service.gov.uk/media/5fce68c18fa8f5788db404f7/NRA_2020_v1.2_FOR_PUBLICATION.pdf - JMLSG (current). Prevention of money laundering/combating terrorist financing: Guidance for the UK financial sector. Sets out the risk-based approach to CDD, EDD triggers, simplified due diligence, and PEP handling in practice. https://jmlsg.org.uk/guidance/ - Signicat (2022). The Battle to Onboard 2022: The growing power of digital identity. Survey finding that 68% of UK and European consumers abandoned a financial services application due to onboarding friction, with verification complexity a key factor. https://www.signicat.com/resources/reports/battle-to-onboard-2022 - FCA (2019). Final Notice to Standard Chartered Bank. £102.2m regulatory fine for financial crime control failures including inadequate customer due diligence and ongoing monitoring. https://www.fca.org.uk/publication/final-notices/standard-chartered-2019.pdf - ICO (current). Guide to the UK General Data Protection Regulation (UK GDPR). Governs data collection, privacy notice requirements, and Article 22 constraints on solely automated decisions in client onboarding contexts. https://ico.org.uk/for-organisations/guide-to-data-protection/guide-to-the-general-data-protection-regulation-gdpr/ - ICO (current). AI and data protection guidance. Sets out transparency, impact assessment, and human review requirements for AI used in risk assessment and onboarding workflows. https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/artificial-intelligence/ - NCSC (current). Principles for the security of machine learning. Recommends logging, monitoring, and access control for AI systems processing sensitive data, directly relevant to AI-assisted onboarding tools. https://www.ncsc.gov.uk/collection/machine-learning - FCA and Bank of England (2019). Machine learning in UK financial services. Documents regulatory expectations for AI governance, oversight, and explainability applicable to firms using AI in KYC controls. https://www.fca.org.uk/publication/research/machine-learning-in-uk-financial-services.pdf

Frequently asked questions

Does my firm need to conduct KYC if we're not a bank or financial services company?

Whether your firm needs formal KYC depends on whether you fall within the sectors covered by the Money Laundering Regulations 2017. Accountants, tax advisers, estate agents, insolvency practitioners, auditors, and company service providers are all in scope. Firms outside those sectors are not legally required to conduct customer due diligence under MLR 2017, though clients in regulated sectors may still require you to complete their due diligence processes as a supplier condition.

What is the minimum a services firm needs to ask clients at onboarding for KYC compliance?

For standard-risk individual clients, the minimum covers full legal name, date of birth, residential address, and preferred contact details, verified against an independent source such as a passport or driving licence. For corporate clients, add the company number, registered address, and confirmation of who holds more than 25% ownership or control. The FCA Financial Crime Guide and JMLSG guidance both indicate this is sufficient for lower-risk clients, with enhanced questions reserved for higher-risk situations.

Can I use AI to automate KYC onboarding, and what are the UK rules I need to follow?

Yes, the FCA has confirmed that firms may use AI in financial crime controls including KYC onboarding, but they remain fully responsible for outcomes. Your AI tool must be explainable, governed, and subject to human oversight. The ICO's guidance on automated decisions under UK GDPR Article 22 adds that solely algorithmic decisions with significant effects on individuals generally require explicit consent or a legal basis plus a human review route. Log all AI interactions and retain records for five years under the MLR 2017.

This post is general information and education only, not legal, regulatory, financial, or other professional advice. Regulations evolve, fee benchmarks shift, and every situation is different, so please take qualified professional advice before acting on anything you read here. See the Terms of Use for the full position.

Ready to talk it through?

Book a free 30 minute conversation. No pitch, no pressure, just a useful chat about where AI fits in your business.

Book a conversation

Related reading

If any of this sounds familiar, let's talk.

The next step is a conversation. No pitch, no pressure. Just an honest discussion about where you are and whether I can help.

Book a conversation