AI-native, with a
human in the loop.

Lebon Consulting (Pty) Ltd (“Lebon”, “we”, “us”) is an authorised Financial Services Provider (FSP 52013) and a registered credit-industry participant (NCRCP21271). This Policy sets out how Lebon uses artificial intelligence (“AI”) across its platforms and services, the disclosures we make to clients, and the controls that govern that use.

Lebon is an AI-native business. AI is not an add-on or a convenience feature; it is a core part of how we deliver financial services efficiently, consistently and at scale. We have deliberately built our platforms around AI so that clients can understand, compare and transact in financial products without the cost and delay of traditional intermediation. We disclose this openly because clients are entitled to know how our services are produced.

At the same time, AI does not displace human responsibility or professional judgement at Lebon. Every material decision affecting a client sits behind a human accountable person, and our AI is constrained to providing factual product information rather than financial advice. This Policy explains both halves of that position.

This Policy applies to all use of AI by Lebon, including but not limited to the following platforms and channels:

The Policy binds all Lebon employees, Key Individuals, representatives, contractors and agents. Where a platform is operated for or with a partner (for example an insurer, institution or funder), this Policy governs Lebon’s own AI conduct; partner obligations are dealt with in the relevant agreement.

Lebon applies the following principles to every AI system it deploys. These principles are consistent with the expectations articulated by the FSCA and the Prudential Authority in their November 2025 joint report on AI in the financial sector.

This section is the legal heart of the Policy. It explains why Lebon’s AI-driven, self-service model is permissible under the Financial Advisory and Intermediary Services Act, 2002 (“the FAIS Act”), and the controls that keep us on the correct side of the advice boundary.

Advice is not compulsory

The FAIS Act separates financial services into two distinct pillars — advice and intermediary services — and an FSP may be authorised for one or both. Lebon is authorised for both, which means we are fully entitled to offer intermediary services only on a given platform, without rendering advice. Because no advice is given on our execution-only channels, Lebon is not required to perform a comprehensive needs analysis or to issue a Record of Advice for those transactions. The client selects the product on their own initiative and we act on that instruction.

What our platforms do — and do not — do

No automated advice

Lebon does not provide automated advice on its execution-only channels. We specifically avoid “guided selling” — a questionnaire or algorithm that converts a client’s answers into a recommendation to buy a particular product. The FSCA treats such guided steering as automated advice, which would attract the full advice-process duties and additional Fit and Proper governance obligations for the underlying algorithm. Where Lebon does at any point elect to offer advice (automated or human-assisted) through a platform, that channel will be designed and governed as an advice channel — with a needs analysis, suitability assessment, Record of Advice and algorithm governance — and will be clearly separated from the execution-only journey.

The execution-only acknowledgement

Before a client submits any application on an execution-only basis, the platform presents a prominent, mandatory acknowledgement (a tick-box that cannot be pre-ticked) confirming that:

• the client is purchasing the product on an execution-only basis;
• no financial advice has been provided by Lebon or its AI;
• any information shown was factual and general in nature; and
• the client accepts responsibility for satisfying themselves that the product meets their own needs, and may consult an independent financial adviser before transacting.

The acknowledgement, the version of information shown, and the timestamp are logged and retained as part of the transaction record.

Statutory disclosures we still owe (Section 13 and General Code)

Operating without advice does not switch off our intermediary-service duties. On every relevant channel Lebon discloses:

AI at Lebon operates under human oversight at all times. The regulator’s expectation — that financial decisions, whether made by AI or by people, must deliver fair outcomes backed by human oversight and governance — is built into how we work.

The following sets the baseline. It is illustrative, not exhaustive; anything not clearly permitted is referred to the Compliance Officer before deployment.

Lebon makes the following disclosures so that clients understand they are dealing with an AI-native provider and know the limits of what our AI does.

The standard customer-facing disclosure text is set out below and is deployed consistently across platforms.

All personal information processed by AI systems is handled in accordance with the Protection of Personal Information Act, 2013 (“POPIA”) and Lebon’s Data Governance Policy. In particular:

• Processing is limited to identified, lawful purposes, with a valid lawful basis and, where required, consent.
• Digital onboarding and any direct marketing comply with POPIA, including the direct-marketing provisions.
• Data minimisation applies — AI systems receive only the personal information necessary for the task.
• Special-category and children's information receive heightened protection.
• Data-subject rights (access, correction, objection, and rights relating to automated decision-making) are honoured.
• We align with guidance issued by the Information Regulator as the FSCA, PA and Information Regulator coordinate on AI and privacy.

Lebon takes reasonable steps to ensure that AI outputs are accurate, current and fair:

• Factual product data is sourced from authoritative inputs and reviewed for currency.
• AI systems are tested before deployment and monitored thereafter for accuracy and unfair bias.
• We distinguish between authoritative product terms (the supplier's documents) and convenience summaries (which AI may generate).
• Identified errors or biased outcomes are logged, remediated and, where they affected a client, put right.

Lebon uses third-party AI services (including large language models) as part of its technology stack across its GCP and Azure environments. Use of these providers is governed by our Outsourcing and Third-Party Risk Policy and the FSCA’s expectations on third-party service provision. We:

• Contract only with providers offering appropriate security, availability and data-protection terms.
• Ensure data-processing terms are in place before personal information is processed by a provider.
• Retain accountability for outsourced AI functions — outsourcing does not transfer our regulatory responsibility.
• Monitor provider performance and maintain contingency arrangements.

Consistent with the FSCA/PA emphasis on governance, model-risk management and board-level oversight, Lebon maintains proportionate controls over its AI models:

Given Lebon’s current scale as a founder-led FSP, these controls are applied proportionately. As the business grows, the governance forum and oversight cadence will be formalised accordingly.

This Policy is designed to align with the existing financial-sector framework (the FAIS Act and General Code of Conduct, the National Credit Act, and POPIA) and with the regulators’ stated direction of travel on AI. South Africa has no dedicated AI statute yet; in the interim, Lebon will revise this Policy as the following instruments are finalised:

The text below is published on Lebon platforms (website footer, chatbot welcome, and the execution-only checkout). It is plain-language and intended for clients.