PRAGMADiscover your offer
Articles

AI in banking, wealth management and family offices in Monaco: what it can prepare, what stays human

Painting: a private banker meets a client, one hand on the dial of a vault that keeps a golden constellation inside

What can AI do for a bank, management company or family office in Monaco?

In a bank, a management company or a family office, AI is mostly useful for preparation: checking that an onboarding file is complete, summarising the history before a meeting, assembling a draft report, tracking official texts. A person still decides. In Monaco, Law No. 1.362 leaves responsibility for due diligence with the professional, even when a third party is used,1 Article 308 of the Criminal Code punishes the disclosure of a secret,2 and Law No. 1.565 protects everyone, subject to exceptions, from a significant decision based solely on automated processing.3

Key points

  • Financial and insurance activities are Monaco's largest sector: 17.4% of 2025 GDP, or 1.8 billion euros.4
  • Banks, firms carrying on a financial activity and multi family offices are expressly covered by anti-money laundering Law No. 1.362.1
  • For these professionals, monitoring of atypical transactions must in principle already be automated; generative AI adds to that system and does not replace it.5
  • Information sent to the AMSF cannot be disclosed to the client or to third parties, apart from exchanges strictly governed by the law, so an AI tool has to be partitioned accordingly.1
  • Assessing an individual's creditworthiness is a high-risk use under the AI Act, which can reach a Monaco company when the output is used in the Union.6

This guide is for executives, compliance officers and portfolio managers in Monaco's financial centre. It starts from the applicable texts, then reviews the main uses worth testing, setting out what the tool prepares and what the person keeps. It does not replace advice from your legal counsel.

A financial centre under a demanding framework

Financial and insurance activities generated the Principality's largest GDP in 2025, 1.8 billion euros, or 17.4% of the total, up from 16.1% in 2016.4 All banks in Monaco, along with fund and portfolio management companies, are members of the Monaco Association for Financial Activities (AMAF), which represents the profession before the authorities.7

The compliance backdrop has just shifted. On 30 October 2026, the government announced that the FATF had removed Monaco from its "grey list" of jurisdictions under increased monitoring, adding that this exit "is not an end in itself".8 The Monegasque Financial Security Authority (AMSF) brings together the financial intelligence unit, supervision and sanctions, and can carry out off-site and on-site inspections.9 It is therefore wise for any AI tool introduced into a due diligence process to be explainable to an inspector.

Which rules apply to whom

The first step is to place your organisation. Several regimes overlap.

Onboarding files and KYC

As a rule before entering a business relationship, the professional identifies the client, any agent and, where applicable, the beneficial owner, and verifies these details against a reliable supporting document.1 For a holding company, trust or foundation, the file can run to many documents in several languages. Chasing documents can take a lot of time here.

The person keeps what matters: assessing the risk, handling a politically exposed person under enhanced measures,1 and deciding whether to enter the relationship. Using a third party does not move the burden: final responsibility for identification and due diligence stays with the professional.1 The same file checks exist in other regulated trades, as our article on AI in Monaco real estate shows.

Monitoring, alerts and suspicion: the most sensitive area

Sovereign Ordinance No. 2.318 already requires banks, financial firms and multi family offices to run a system for monitoring atypical transactions. It must cover all accounts, rest on precise criteria, produce written reports, undergo initial validation and periodic review, and be automated, unless an exemption is justified to the AMSF.5 So the question is not whether to automate, but what a language model adds to that system.

We found no AMSF guideline specific to AI as of this article's date. The closest reference is French and does not address Monaco: an explanatory document, the ACPR and Tracfin guidelines of 23 April 2025 mention alert prioritisation, enriching client knowledge and drafting assistance. They stress that the institution remains responsible for the explainability of its algorithms and that information produced by AI must undergo appropriate and proportionate human control, particularly for enhanced reviews and the drafting of suspicious transaction reports.12

Two Monaco rules mark the boundary. The report to the AMSF is made in writing, as a rule before the transaction is executed, and states the facts on which it rests.1 And, apart from the exchanges governed by Article 45, it is forbidden to bring information sent to the Authority to the attention of the client or a third party.1 In practice, suspicion files stay out of reach of the assistant used by client-facing teams: otherwise a meeting note could reveal what must remain confidential.

Meetings, reporting and regulatory watch

These uses carry the least exposure and often make the best first trials, provided the sources are under control.

Gains are measured on your own work. In a trial across 66 firms, staff using an assistant built into their tools spent two fewer hours a week on email;13 in Denmark, another study nevertheless detects no effect on earnings or recorded hours.14 Saved time only becomes value when the work is reorganised around it. Professionals juggling expiring documents and international clients will recognise the same logic in our article on AI in Monaco yachting, and our guide to AI use cases in Monaco sets it within the Principality's economy.

Professional secrecy: choose the tool before the use case

Article 308 of the Criminal Code punishes with six months to one year in prison and a fine, or either penalty, anyone entrusted with a secret by status or profession who reveals it outside the cases allowed by law.2 Law No. 1.338 expressly applies this to people who run or work for a licensed company; the secret cannot be raised against the CCAF or against the judicial authority acting in criminal proceedings.10

Sending a client file to an external AI service is therefore a legal decision before it is a technical one. Have your counsel settle whether that disclosure is compatible with secrecy, then check at least the following.

The sector is concerned: in the Bank of England and FCA's 2026 survey, generative AI makes up 56% of reported AI systems, and the highest-rated risks are data protection and security, third-party and supply chain risk, and risks specific to generative AI, including hallucinations.16

Automated decisions and the AI Act

Everyone has the right not to be subject to a decision producing legal effects or significantly affecting them that is based solely on automated processing, including profiling. Except where the decision is authorised by a legal text, which then sets its own safeguards, an exception leaves the person at least the right to obtain human intervention, express their point of view and contest the decision.3 Refusing a relationship or a loan on a model's score alone should therefore be ruled out. A systematic and extensive evaluation of personal aspects underpinning such decisions also calls for a prior impact assessment.3

Monaco is not an EU member, but the AI Act can reach a Monaco company that places an AI system on the EU market, or whose system output is used in the Union.6 This is worth checking for an entity serving clients or a parent company in the Union. Systems intended to assess the creditworthiness or credit score of individuals are classed as high risk, except for detecting financial fraud;6 Regulation (EU) 2026/1744 sets the application of their obligations at 2 December 2027; a system already placed on the market or put into service before that date is, as a rule, caught only if its design changes significantly.17 Our article on the AI Act for Monaco companies sets out these criteria.

Where to start

Pick a frequent task with no decision about a person that is easy to check: a file completeness review or a meeting preparation note. Test it on fictitious or pseudonymised files, have compliance approve the scope, then measure the actual time and the errors your teams find.

On infrastructure, PRAGMA offers European hosting, or hosting in Monaco on request, subject to technical and contractual validation. Our page on AI for family offices and firms in Monaco describes this approach.

Frequently asked questions

Can a Monaco bank use AI for KYC?

No text we reviewed forbids it, but responsibility for due diligence stays with the professional, even when a third party carries it out.1 AI can check completeness and prepare the analysis; the decision to enter the relationship stays human.

Is a family office subject to anti-money laundering law in Monaco?

Law No. 1.362 expressly covers multi family offices.1 A structure dedicated to a single family should have its position checked, particularly according to the activities it carries on.

Can we send a client file to ChatGPT or another public assistant?

Not without prior analysis. The file may fall under professional secrecy2 and contain personal data, which requires, if the provider acts as a processor, a written contract and sufficient guarantees and, outside adequate countries, a legal basis for transfer.3 Use a contractually covered environment and remove identifiers whenever possible.

To find out which task in your firm can be tested without touching a decision or client secrecy, start with our free AI audit.

Request the free audit and prototype

Sources

  1. Légimonaco, Loi n° 1.362 du 3 août 2009 relative à la lutte contre le blanchiment de capitaux, le financement du terrorisme et de la prolifération des armes de destruction massive et la corruption,
  2. Légimonaco, Code pénal de Monaco, article 308 (rédaction de la loi n° 1.444 du 19 décembre 2016),
  3. Légimonaco, Loi n° 1.565 du 3 décembre 2024 relative à la protection des données personnelles,
  4. IMSEE, Produit Intérieur Brut 2025,
  5. Légimonaco, Ordonnance Souveraine n° 2.318 du 3 août 2009 fixant les conditions d’application de la loi n° 1.362,
  6. EUR-Lex, Règlement (UE) 2024/1689 sur l’intelligence artificielle,
  7. Gouvernement Princier, Mon Entreprise, Monaco Association for Financial Activities,
  8. Gouvernement Princier, Monaco Exits the FATF “Grey List”, the Global Anti-Money Laundering Watchlist,
  9. Gouvernement Princier, Combating money laundering and the financing of terrorism,
  10. Légimonaco, Loi n° 1.338 du 7 septembre 2007 sur les activités financières,
  11. Légimonaco, Accord du 3 décembre 2018 par échange de lettres entre la France et Monaco relatif à la réglementation bancaire applicable dans la Principauté,
  12. ACPR et Tracfin, Lignes directrices conjointes relatives aux obligations de vigilance sur les opérations et aux obligations de déclaration et d’information à Tracfin,
  13. NBER, Shifting Work Patterns with Generative AI, Working Paper 33795 (Dillon, Jaffe, Immorlica, Stanton),
  14. NBER, Still Waters, Rapid Currents: Early Labor Market Transformation under Generative AI, Working Paper 33777 (Humlum, Vestergaard),
  15. APDP, Publication de la liste des pays, territoires et organisations internationales disposant du niveau de protection adéquat,
  16. Bank of England et FCA, Artificial intelligence in UK financial services 2026,
  17. EUR-Lex, Regulation (EU) 2026/1744 (Digital Omnibus on AI),

Add PRAGMA as a preferred source on Google