Artificial intelligence (AI) is rapidly transforming the financial services landscape in Mauritius, offering new opportunities for efficiency, innovation, and customer engagement. Recognising both its potential and the risks, the Financial Services Commission (FSC) has released its Fintech Series Guidance Notes No. 4 on the Responsible Use of AI in Financial Services in September 2025 (Guidance Notes). The Guidance Notes apply to the sectors of insurance, wealth management and non-banking financial institutions (NBFIs) subject to FSC oversight, covering both traditional AI and generative AI.
Use of AI across the global financial sector
Financial institutions across the global financial sector are leveraging AI in various areas, including robo-advisory services, portfolio optimisation, natural language processing for market analysis, algorithmic trading, insurance pricing and claims analytics, anti-money laundering (AML) and fraud detection, risk modeling, and customer service chatbots and virtual assistants.
Principles for responsible use of AI
The Guidance Notes present a unified framework for responsible AI built on four mutually reinforcing pillars:
· Governance: Strong governance is the foundation, requiring boards and senior management to maintain clear oversight and accountability throughout the AI lifecycle from design and approval to deployment, monitoring, and decommissioning with defined roles, adequate AI literacy across teams, and human oversight where consumer outcomes may be materially affected.
· Fairness and Bias Mitigation: Fairness and bias mitigation are integral to this governance model. Institutions should design and manage systems to avoid harmful or discriminatory outcomes, foster an ethical culture supported by regular training and cross‑functional teams, and implement continuous monitoring and auditing using techniques such as adversarial testing and anomaly detection to ensure explainability and transparent decision making.
· Transparency: Transparency, in turn, enables trust and fair treatment by ensuring customers receive clear, timely, and adequate information while balancing confidentiality, intellectual property protection, and legal obligations, particularly in sensitive areas like fraud detection. Firms should maintain robust documentation and assurances for third‑party systems and provide effective channels for engagement, information requests, and complaints.
· Security: Security underpins and operationalises these commitments through ongoing validation, robustness testing, and automated monitoring to detect data drift and trigger recalibration, coupled with up‑to‑date cybersecurity controls and regular staff training; where third‑party AI is used, significant robustness findings should be shared to support corrective action. Together, these elements form a coherent, end‑to‑end approach that embeds ethical, transparent, and resilient AI practices across the institution and its vendors, ensuring accountable use and strong consumer protection.
Additionally, in order to guide its licensees seeking to develop, deploy and use AI technologies, the FSC has developed a set of “Principles for the responsible use of AI” which consist of nine principles: fairness and bias mitigation, transparency, accountability, privacy, security, environmental sustainability, human-centricity, continuous monitoring and evaluation and compliance ethics.
Potential benefits and risks applicable to the adoption of AI
The FSC emphasises the potential benefits of AI in the financial sector which include: improved regulatory compliance, increased revenue and value creation, enhanced decision-making, greater financial inclusion, operational efficiency, better consumer experiences, more accurate investment forecasting, stronger cyber resilience, and ongoing staff development.
Nevertheless, the risks and challenges associated with the adoption of AI should not be overlooked. These include bias and discrimination, privacy and data protection concerns, model opacity, systemic and concentration risks, cybersecurity threats, outdated legacy systems, skills shortages, budget constraints, and limited access to platforms and tools.
Data protection considerations
The Guidance Notes reinforces the statutory obligations set out under the Data Protection Act 2017 (DPA)_, particularly for automated decision-making and profiling that produce legal or significant effects for individuals. Accordingly, organisations must ensure strict compliance with the applicable provisions of the DPA (including informing data subjects of such automated decision-making processes and conducting the relevant data protection impact assessments to identify, evaluate, and mitigate risks).
As AI continues to evolve, the Guidance Notes provide a solid framework for responsible innovation, helping Mauritius’ financial sector harness the benefits of AI while protecting consumers and upholding the highest standards of ethics and compliance.
