WealthAi launches AI agents for wealth onboarding and KYC

WealthAi, a London-based wealth management technology provider, says it launched a suite of AI agents on 2 October 2026 designed to run client onboarding and know-your-customer (KYC) checks from first contact through to account opening. The company's own announcement describes agents that collect documents, spot missing information, generate forms and set in motion identity verification, politically exposed person (PEP) screening, sanctions screening, KYC checks and account verification.

The launch matters to UK wealth managers because onboarding is one of the most document-heavy, compliance-sensitive parts of client acquisition, and because the firm doing the onboarding — not the technology vendor — carries the regulatory responsibility. WealthAi's own Trust Centre material says exactly that: the company states it is not regulated by the Financial Conduct Authority (FCA), and that client firms remain the regulated entity, the data controller, and the decision-maker for client-facing outcomes. That distinction, between a system that coordinates a workflow and a firm that must still make and own regulated decisions, is the question any adopting wealth manager needs to work through before switching the agents on for real clients.

What the agents are said to do

According to WealthAi's launch material, the agents act as coordinators across clients, advisers, compliance teams, operations staff and the firm's other systems, rather than simply automating one isolated check. The stated workflow runs roughly as follows: an agent gathers client information and documents, flags what is missing, generates the relevant forms, and then initiates identity verification, PEP and sanctions screening, KYC checks and account verification.

WealthAi's product material says specialist identity-verification and compliance services can be connected at the relevant stages of that workflow, rather than WealthAi performing every underlying check itself. An earlier indexed version of the launch page specifically named Tiller Technologies as that specialist partner; the version currently accessible does not name Tiller or any other provider. Readers should treat the identity of the verification partner, and whether Tiller remains part of the live capability, as unconfirmed until WealthAi or Tiller states otherwise.

WealthAi says people are brought back into the process "where review or approval is required." The launch announcement does not set out the rules that trigger that review, the exception thresholds, how false positives are handled, or who holds the authority to approve or reject a case. WealthAi's Trust Centre adds that sensitive outcomes can be routed to human review and that the system records the reviewer, the time and the decision taken, as part of what it describes as a chain of activity covering model outputs, reviewer actions and follow-up.

Orchestration is not the same as a regulatory decision

It is worth separating two things that this kind of launch can blur together: moving a case through a workflow, and making the judgement that UK rules either require a senior person to make or that a firm routes to a person for meaningful human challenge.

The Money Laundering Regulations 2017, in force since 26 June 2017, require relevant firms to carry out customer due diligence and to maintain risk-management systems capable of identifying whether a customer or beneficial owner is a politically exposed person, a family member, or a known close associate. A screening result — a name match, a risk score, a flag — is an input to that risk assessment. It is not, by itself, the regulatory decision.

Three points in the workflow stand out as places where this distinction matters most:

PEP relationships. Regulation 35 of the 2017 Regulations requires senior-management approval before a firm enters into or continues a business relationship with a PEP, alongside adequate measures to establish source of wealth and source of funds, and enhanced ongoing monitoring. FCA guidance published on 7 July 2025 (FG25/3) says firms should apply a proportionate, risk-based approach to UK PEPs, their relatives and close associates — the starting point for a domestic PEP is lower risk than for a non-domestic PEP where no enhanced risk factors are present, but the approval requirement itself is not something a firm can delegate to an autonomous agent.

Sanctions ambiguity. Since 28 January 2026, the UK Sanctions List has been the sole source for UK sanctions designations, according to Office of Financial Sanctions Implementation (OFSI) guidance. But government guidance is clear that sanctions due diligence cannot be reduced to an exact-name match against that list: a firm also has to consider whether an unlisted entity is owned or controlled by a designated person. A firm needs an effective escalation path for that analysis, with meaningful, overturn-capable human review wherever a case is routed to a person, rather than treating a clean list match as the end of the check.

Unresolved identity evidence. Where a client's identity evidence does not clear automated checks cleanly, a firm needs an escalation path or an alternative verification route rather than treating the automated result as final. Separately, HM Treasury and the Department for Science, Innovation and Technology said on 26 February 2026 that any digital verification service used for identity checks under the Money Laundering Regulations must appear on the GOV.UK register of services certified against the UK digital verification services trust framework. That guidance supplements, rather than replaces, the underlying regulations: a firm has to ensure the verification service it relies on is itself registered and certified, and routing an unregistered service's output to a human reviewer does not make that use compliant.

Accountability sits with the wealth manager

Where an FCA-authorised firm outsources a critical or important operational function, the firm remains fully responsible for it; outsourcing rules do not let a firm transfer senior personnel's accountability or change its obligations to clients. WealthAi's own position — that it is not FCA-regulated and that the client firm is the regulated entity — is consistent with that framework, but it means the onboarding firm needs to work out, case by case, whether its use of these agents amounts to outsourcing a critical or important function, and what due diligence, contractual controls and exit arrangements that requires.

Auditability and data protection

WealthAi's Trust Centre claims the platform retains an auditable chain covering model outputs, reviewer actions, follow-up and system activity, and that reviewers, times and decisions are logged. These are vendor claims rather than independently assured facts, and no public independent assurance report testing them was identified as of 4 October 2026.

The Information Commissioner's Office (ICO) says that where AI processing is likely to create a high risk to people's rights and freedoms, a data protection impact assessment should document the processing, the data flows, error margins, the controller and processor roles, and where meaningful human involvement occurs. Critically, the ICO says that where a human reviews an automated outcome, that review has to be meaningful: the reviewer must be able to overturn the decision, not simply wave it through. A firm adopting these agents needs to be able to show, not just assert, that its reviewers have the evidence, authority and time to do that, and that the arrangement is not a rubber stamp under volume pressure.

The Consumer Duty test

The Consumer Duty, which began applying to open products and services from 31 July 2023, covers prospective retail customers as well as existing ones. That makes onboarding delays, inaccessible evidence requests, unexplained rejections and weak escalation routes genuine Consumer Duty issues, not just operational friction. A firm considering these agents should be able to show its governing body evidence on completion times, abandonment, repeated evidence requests, how false positives are resolved, and whether outcomes differ for customers with characteristics of vulnerability — none of which is addressed in WealthAi's own launch material.

What is still unverified

WealthAi's launch headline refers to onboarding happening in "a fraction of the usual time," but the announcement supplies no baseline, measured reduction, sample size or independent validation for that claim. Separately, the launch page's indexed and currently accessible versions disagree on whether Tiller Technologies is the integrated verification partner, and WealthAi's page metadata has shown a launch date inconsistency with the 2 October 2026 dateline used in this article. None of these points has been resolved on the public record as of 4 October 2026.

What to watch next

Wealth managers considering this capability can check the FCA's SYSC 8 outsourcing rules and its Consumer Duty rules in the Handbook, HM Treasury's February 2026 guidance on digital identity services, OFSI's sanctions guidance, and the ICO's published material on AI accountability and data protection impact assessments, all of which set out the obligations a firm keeps regardless of which technology sits behind its onboarding process. Confirmation from WealthAi or Tiller on the verification-partner question, and any independently measured onboarding or accuracy figures, would also help firms assess the claim on its merits rather than on the launch copy alone.