AMAF: “AI and finance, how businesses will have to reinvent themselves”

Faced with Monaco's inclusion on the FATF Grey List, the Monaco financial marketplace, which currently has nearly €160 billion in assets*, must manage the development of artificial intelligence (AI).  Within the Monégasque Association of Financial Activities (AMAF), a working group is currently looking at the issue.   Whilst according to a Citigroup study, AI could replace around 54% of jobs in the banking sector, due to the strong automation potential, AMAF President, Robert Laure, does “not imagine any shock" to employment.

Has the Monégasque banking sector been impacted by Monaco's inclusion on the FATF Grey List?
Maintaining enhanced surveillance and the Principality's subsequent inclusion on the Grey List has not been not without effect on banking activities.  However, in terms of the quantitative aspect, we have not observed any shock phenomenon.  At the end of 2023, the outstanding assets held by local banks amounted to nearly €160 billion*.  It was estimated at the end of September, three months after the FATF's decision, at €168 billion, a positive development that includes an estimated stock market revaluation effect of these assets, combined with a quasi-stable flow effect linked to its common volatility.

Was there no deterrent effect?
No - and this is also thanks to the communication work carried out in advance of the FATF decision by our members.  They spoke extensively to their clients and the banks, which all belong to international groups, all spoke to their parent company.  These same entities were wondering about a situation in contrast with the observations of the Moneyval report, which highlighted the preparation and compliance of the Principality's financial institutions with LBT/FT-C standards and rules.  However, in its communication, the FATF stressed the need for the Principality to be equipped with appropriate means for the proper application of the anti-money laundering system.  A six-point action plan and an 18-month timetable were given for this to be achieved.

In real terms, how does being on the Grey List unsettle financial institutions?
There are two main unsettling points.  First of all in the relationships that our financial institutions have with their partners outside the Principality who, because our inclusion on the Grey List, can and often do increase their requirements in terms of requests for additional information.  This results in an administrative burden and an increase in the costs of operational processing for local institutions.
Next, in an exceptionally short period of time, the Principality adopted a legal and regulatory corpus that sometimes over-transposed international standards.  This system interacts with the long-established one already in place.  A period of appropriation and adaptation is therefore essential.

Has the war in Ukraine had no impact, even in relation to the capital of Russians or the Ukrainians under sanction?
Currently the non-resident Russian clientèle represents less than 1% of the volume of the Monaco market's outstanding sums.

Some Russians have a dual European nationality, Maltese or Cypriot, Hungarian….
There was a period when certain countries of the European Union granted so-called ‘‘golden passports’’.  These passports, granted quasi-automatically in return for financial and/or economic commitment to the host country, are no longer really current.  Monaco has never been concerned by this phenomenon.

Do you think AI applied to the management of portfolios has had a positive effect on funds managed in Monaco? Is this a subject of study of the AMAF?
There is a lot of imagination surrounding artificial intelligence, presented as the alpha and omega of portfolio management and which would be capable of making sure-fire winning decisions in financial investments.  However, portfolio management remains a combination of science and art.  It is a science in that the processing of data, an essential source of portfolio management, is disrupted with the arrival of artificial intelligence tools.  It is an art in that the rationality of investment choices is still influenced by human emotion.
This year, the AMAF created the Digital Affairs Working Group, which aims to address all digital activities that have or may have an impact on our financial activities.  This therefore concerns both technological developments and regulatory developments (blockchain, AI, cryptos).

What are the priorities of this working group?
First of all, to ensure technical and regulatory monitoring in the digital field for financial activities; then to assess the operational applications of these devices for financial institutions in the market - for example, the European DORA regulation which concerns elements related to the security of IT and communication service providers.  Finally, it seeks to identify the risks posed by the profound digital transformation of financial services, the growing interconnection of networks and critical infrastructures, as well as the multiplication of increasingly sophisticated cyberattacks against players in the financial sector.

What is the concrete contribution of AI in terms of financial activities?
The ability to process data of almost unlimited dimensions and in extremely short time frames has been made possible by technological development.  This is the first revolution.  Historically, banks and insurance companies were the first to use the computer as a 'machine tool' for processing the mass of information to be exploited.  The contribution of AI in this data processing allows the development and sophistication of risk management, customer file management, such as Know Your Customer (KYC), or even the understanding of customer behaviour.  The other development path, generative AI, is also found in office automation.  It helps with writing, facilitates the search for information, makes it easy to translate into languages that are sometimes little-used, allows the immediate writing of meeting minutes - in short, ideal assistance.  It is instantaneous, it is not always perfect but it saves time.  A recent study by McKinsey estimated productivity gains for companies that use generative artificial intelligence at 20 to 30%.

 Will artificial intelligence now be able to make the best possible investment choices?
That is to say, will artificial intelligence beat biological intelligence in terms of investment choices? It's a great debate.  This reminds me of the Dart Portfolio of the 1970s and 1980s, according to which a chimpanzee throwing darts at a list of stocks built a stock portfolio that performed as well, if not better, than that of a professional manager.  In those years, the exploitation and dissemination of information were much less advanced than today! Now, the mastery and exploitation of information by professional managers is essential in order to manage effectively.
For me, there are three intangible rules for financial investment and wealth management: work; discipline, ie emotion management; and time (when you make a financial investment, all you are doing is deferring a future consumption).
Artificial intelligence has a competitive advantage.  It has this ability to forget nothing, to lose nothing and to be able to analyse everything.  In terms of work, it masters all the information.  In terms of discipline, if it is well-educated, if we give it the right rules, it will apply them without emotion.  Finally, time is not a biological concept for computers.  The computer is there for eternity.
In my opinion, artificial intelligence will be and will remain an extremely powerful and useful tool at the service of biological intelligence.  The machine will accompany man; but it is man who will have to transmit knowledge to the machine.  AI will save time, increase efficiency and productivity and, therefore, the development of activities.
Establishing the best possible choices in terms of investment will surely be associated with a more common and advanced use of AI.

Including for credit management?
Of course.  Today, banks and finance companies are applying increasingly precise scoring on the loans to be granted.  A particular point of attention is the control of the supply of information to AI applications and the monitoring of the evolution of the learning capacity of AI.  Risks of behavioural bias have already been identified.  Whether in terms of gender, social or behavioural position, people may find themselves excluded a priori from access to credit.  The intervention of human decision must remain a priority.
In fact, the adoption of a European regulation on artificial intelligence, the AI Act, aims to avoid these risks, amongst other things.

Will it be transposed here? Monaco is part of the Eurozone?
By application of the monetary convention with France and now of the European Union, Monaco transposes the directives and applies the regulations in the field of banking activities.  However, the scope of the AI Act is broader.  Its transposition, in connection with banking activities, will, therefore, follow a national legislative process.  The implementation of the AI Act is planned for 2026 in member countries.  Monaco should not remain outside this development.

What are the objectives of the AI working group for 2025?
They will be ambitious - the subject requires it - but above all pragmatic.  The working group has already identified concrete work to improve digital uses with the administration.  The need to use and disseminate cybersecurity tools to all our members will also be an important objective.  However, the working group remains a force for proposals and will present its 2025 work programme to the AMAF at the beginning of the year.

Employment is one of the fears linked to AI...
Financial activities in the Principality, banks and management companies employ around 3,300 employees and this figure is increasing regularly.  Artificial intelligence will inevitably have an impact on employment but from a quantitative point of view, it's tricky to estimate.  Jobs and functions will have to evolve.  I don't imagine a shock but perhaps a faster evolution than some would like: less than a generation seems likely.

People have time to train over a generation.  Are you going to add training in AI?
That is the whole point of the current training provided to the profession, including those set up at the initiative of AMAF.  They are legally mandatory for financial banking activity, anti-money laundering and information on risk control.  Only those holding the certification are authorised to practice these jobs.  Training in AI is sometimes mentioned.  However, jobs related to IT in the traditional sense are set to evolve significantly or even disappear.  The job of coder would be on this list.
The use of artificial intelligence requires mastery of fundamental knowledge, such as reading, writing and calculation.  If you are unable to clearly formulate your needs or requests, AI may provide unsuitable answers.  Optimal use of AI, therefore, relies on a solid capacity for analysis and reflection, essential prerequisites to guarantee effectiveness.