“Those who attack or dismiss AI are scared of what they do not fully understand”

Noor Yadallee, Founder of dot42.dev

AI holds immense potential for the Mauritian economy. For Noor Yadallee, Google Developer Expert and Founder of dot42.dev, the benefits of buying into the AI revolution for small island developing states like Mauritius are wide-ranging. To realise its full potential, we must first establish robust data governance and clear regulatory frameworks. Strategic investments will also have to be targeted across our core economic pillars. Although he believes we must not become passively dependent on AI and humans must always provide critical oversight, he says that we should focus on shaping how the technology is used and governed, instead of fearing that it might replace us.

S.K.

Nowadays, the world’s greatest powers use AI to modernise, innovate, and achieve better productivity in their economies. Shouldn’t this also be the case for Mauritius?

We are at a profound global turning point. Artificial intelligence is fundamentally resetting what nations can achieve, decoupling economic scale from sheer population size and redrawing the maps of global productivity. Historically, Mauritius has lagged a decade or more behind the bleeding edge of technological innovation, absorbing advancements only after they became commoditised. Today, however, AI presents an unprecedented, high-leverage transition moment where the playing field has been somewhat flattened. The structural barriers that once constrained small island developing states such as geographic isolation, limited physical landmass, and a restricted domestic labour pool are rendered far less prohibitive in a digital economy powered by autonomous systems, predictive analytics, and generative models.

Realising this potential requires a fundamental shift in trust: moving away from legacy institutions set in their ways and taking a leap of faith towards high-agency youth and local startups. These emerging innovators are actively learning the cutting edge, moving at market speed, and operating without the drag of outdated playbooks. However, ideological willingness alone will not build an AI-driven economy. Mauritius must back this vision with targeted strategic investments across its core economic pillars:

• Smarter Financial Services: Upgrading banking sector processes through automated compliance, predictive fraud detection, and multi-language AI agents to move up the value chain from routine back-office processing to high-margin global wealth management and advisory services
• Precision Agriculture & Climate Resilience: Deploying AI-driven spatial data and satellite imagery to optimise crop yields, manage scarce freshwater resources, and monitor ocean ecology against climate volatility.
• Sovereign Infrastructure & Localised LLMs: Investing in localised compute infrastructure, and regional data centers while training localised AI models attuned to dialects (not just Mauritian Creole) and regional legal contexts.
• Public Sector Efficiency: Streamlining bureaucratic bottlenecks through automated public service delivery, intelligent traffic management, and predictive healthcare diagnostics in public facilities.

“The financial and healthcare sectors are at the absolute forefront of AI adoption.”

By modernising regulatory frameworks, such as establishing regulatory sandboxes that allow startups to test AI applications in real time, aggressively re-skilling its workforce through technical education partnerships, and empowering this rising generation of tech-forward builders to lead national transformation, Mauritius can leapfrog entire cycles of development. Converting small-state agility into a true global advantage will allow the island to transition from a technological consumer into a regional innovation hub, securing its place on the world stage.

In the 2026-2027 Budget, the government clearly expressed its intention to use AI, especially in a difficult economic context. What is your opinion?

While aggressive AI adoption is undeniably essential, and the government’s push through fiscal incentives, upskilling initiatives, and high-profile partnerships with global tech giants such as the Memorandum of Understanding between Mauritius Telecom and Amazon Web Services sets a promising trajectory, we must realise that leveraging AI algorithms is not the starting line. In a constrained fiscal environment, jumping straight to unguided AI deployment risks allocating public capital toward volatile technologies that yield diminishing returns. AI models are ultimately only as effective, unbiased, and impactful as the underlying data that feeds them.

Before Mauritius can safely capitalise on AI, we must first establish robust data governance and clear regulatory frameworks. This requires building secure storage infrastructure, standardising data interoperability across public and private sectors, and enforcing strict data privacy standards. Without explicit policy guidelines and institutional oversight, organisations risk falling prey to “shadow AI,” where employees routinely leak sensitive intellectual property or proprietary state data to untrusted commercial models. Furthermore, relying on unverified algorithmic outputs without human-in-the-loop safeguards opens the door to severe operational, legal, and financial losses triggered by model hallucinations and fabricated factual data.

To prevent these pitfalls, technical training and public education cannot be limited to basic digital literacy; they must instil rigorous data hygiene, risk management, and mandatory human oversight protocols. Without centralised open-data frameworks and transparent compliance guidelines, local developers are left building on fragmented, insecure foundations. Realising the vision of a digitally sovereign, AI-powered nation requires us to treat data as our primary national infrastructure, putting cloud connectivity, cybersecurity, and regulatory guardrails on par with roads and power grids. By securing these fundamentals, Mauritius ensures that those building solutions for the island operate within a safe, reliable ecosystem equipped to drive genuine, long-term economic transformation.

In your view, which struggling sectors could use AI to become more efficient?

Artificial intelligence eliminates domain-agnostic friction by serving as an intelligent layer that sits above fragmented software and departmental silos. Across any industry, organisation-wide drag stems from routine administrative overhead: manually translating data between incompatible systems, tracking down cross-departmental context, or waiting on multi-tier signoffs. AI models dissolve these barriers. By automating mundane coordination, document synthesis, and data reconciliation across all business units, AI shifts the focus of the workforce from managing structural friction to driving strategic output.

The sectors I consider best positioned to benefit are knowledge-intensive industries with lots of repetitive, menial work burdened by heavy coordination costs, data fragmented across specialised tools, and strict compliance overhead. Legal and business operations, which rely heavily on document synthesis and multi-step reviews, are top contenders. We are already seeing practical legal AI platforms being deployed to navigate regional jurisprudence and streamline case analysis. High-volume data reconciliation and regulatory compliance in financial and corporate services make them prime candidates as well.

Beyond knowledge work, AI can revitalise industries traditionally labour-constrained and facing economic distress:

• Agro-Industry & Local Food Systems: Precision agriculture can transform this sector by analysing sensor networks and satellite or drone imagery to optimise water, fertiliser, and harvest timing while cutting input costs. These data-driven tactics maximise crop yields, combat climate volatility, and protect national food security without expanding physical land footprints.
• Manufacturing & Export Textiles: Facing severe pressure from foreign competition, rising shipping costs, and persistent labour shortages, local manufacturing can leverage computer-vision quality control, predictive machine maintenance, and AI-driven supply chain planning. Generative tools also allow design teams to rapidly prototype patterns and garments based on real-time global trend analysis, drastically shortening time-to-market and lowering overhead.
• Port Logistics, Supply Chain & Freight: As an island economy heavily dependent on sea freight, delays at the port ripple through the entire domestic market. AI models can optimise container stacking, predict vessel arrival queues, and automate customs clearance paperwork to minimise vessel turnaround times, cut storage fees, and lower overall import costs.

By applying intelligent systems to these structural bottlenecks, various sectors can lower operational costs, recover margins, and improve their competitiveness in global markets.

Which sectors are already using AI and are likely to use it even more?

The financial and healthcare sectors are at the absolute forefront of AI adoption, using it as foundational infrastructure rather than a mere efficiency tool. In financial services, high-speed machine learning models power real-time fraud detection by analysing millions of transactions per second, while sophisticated alternative credit scoring models expand lending access to previously underserved populations. Simultaneously, automated compliance systems process evolving anti-money laundering regulations and cross-reference global sanctions lists instantly to eliminate manual bottlenecks. 

“Before Mauritius can safely capitalise on AI, we must first establish robust data governance and clear regulatory frameworks.”

In medicine, transformational impact is already visible through deep-learning algorithms that analyse medical imaging like X-rays and MRIs with remarkable precision, helping clinicians catch early-stage diseases and tailor drug therapies to individual genetic profiles. 

Beyond these two pillars, sectors like retail, agriculture, manufacturing, and cybersecurity are rapidly scaling their own AI integrations, spanning from predictive supply chain management to autonomous threat response. 

For a developing digital economy like Mauritius, which has successfully leveraged tech innovation in its financial sector, applying that same strategic urgency to healthcare could revolutionise public medicine. Deploying cognitive tools in public hospitals would streamline emergency triage, relieve heavy patient backlogs, and give regional clinics access to expert-level diagnostic accuracy, ultimately driving better health outcomes for the entire nation. 

Leaving aside the sectors contributing directly to the economy, could AI also be useful in areas such as transport?

Sectors that are rarely treated as standalone revenue generators, such as national transport, should still be enhanced using AI just like direct economic engines. When implemented effectively, AI transforms basic public infrastructure into an intelligent, low-friction framework that delivers compounding benefits across the entire macro economy.

In traffic grid management, replacing legacy, fixed-timer traffic signals with AI-driven adaptive traffic control systems allows intersections to analyse real-time camera and sensor feeds, dynamically adjusting green-light intervals based on actual vehicle density. This eliminates bottlenecks along major arteries like the M1 motorway, buying back hundreds of thousands of lost working hours, cutting fuel consumption, and reducing commercial logistics delays. 

In public transit, AI can optimise networks by integrating bus feeder routes with light rail systems like the Metro Express. Predictive demand models allow transport operators to dynamically route buses, adjust frequency during peak rush hours, and launch demand-responsive transit, ensuring higher coverage with lower operational overhead.

AI also shifts infrastructure care from reactive repairs to predictive maintenance. By analysing vehicle sensor logs and telemetry data, authorities can identify road degradation, potholes, or fleet mechanical failures before catastrophic breakdowns occur. These systemic optimisations do far more than clear roads; they lower national fuel import bills, reduce carbon emissions, and lower the cost of moving goods and people across the island.

Similarly, public healthcare operates as a foundational enabler rather than a direct revenue driver, making it another prime candidate for AI transformation. In public health systems, administrative friction, long waiting times, and resource misallocations directly erode economic productivity through extended sick leave and worker absenteeism. Integrating AI into hospital operations through predictive bed management, automated patient triage, and intelligent appointment scheduling minimises waiting room congestion and streamlines doctor workflows. On a clinical level, deploying computer-vision models for AI-assisted diagnostics, such as rapid analysis of radiology scans or early screening for chronic conditions, allows regional health centers to detect severe illnesses faster and with greater accuracy. Catching medical issues early before they require costly tertiary care lightens the burden on public expenditure. 

By optimising both public transit and public healthcare, a nation reduces the hidden operational tax on its society, unlocking higher overall productivity, optimising public spending, and maintaining a healthier, more mobile workforce. 

AI could also reduce the bureaucracy that heavily burdens public services and make life easier for Mauritians, couldn’t it?

Yes, absolutely. For citizens and businesses alike, dealing with public administration often means navigating fragmented websites, long queues, and rigid paper-based approvals. Deploying AI across public services can fundamentally dismantle this bureaucratic friction, replacing slow manual processing with a proactive, citizen-centric design. Instead of waiting for weeks for civil servants to manually check uploaded support documents, document-AI systems can instantly verify identification, proof of address, or tax returns at submission, flagging errors immediately so applications are valid on day one. Applying for environmental, building, or operational permits often stalls in cross-departmental handoffs. AI can route applications dynamically, auto-check compliance against zoning rules, and flag only edge cases for human review, reducing processing times from months to days!

So, should we stop demonising AI?

Those who attack or dismiss AI are often simply scared of what they do not fully understand. Throughout history, major shifts, from the industrial revolution to the advent of the personal computer, have initially faced intense pushback and fearmongering. Yet, history consistently shows that technology itself is fundamentally neutral. At its core, AI is a powerful tool designed to amplify human capability, streamline tedious workflows, and make us far more efficient in our daily lives, and it is what we do with it that makes it such a polarising tool.

Adapting to this shift will certainly require proactive upskilling, but the barrier to entry has never been lower. Learning resources, courses, and accessible interfaces are widely available, allowing anyone with internet access to adapt and build digital literacy. It is also a common misconception that AI will outright replace humans. In reality, AI does not replace you; it is the person who learns how to leverage AI who will replace the person who refuses to adapt. By automating routine tasks, AI actually frees us up to focus on higher-level problem solving, strategic thinking, and creative innovation.

At the same time, rejecting the “demonisation” of AI does not mean accepting it blindly. Maintaining a balanced perspective is essential. We must avoid becoming passively dependent on these systems and ensure we always keep humans in the loop to provide judgement, ethics, and critical oversight. Ultimately, instead of fearing AI, we should focus on shaping how it is used, governing it responsibly while using it to elevate human potential.

How can we ensure that AI does not become the perfect replacement for humans? 

A 1979 IBM training manual said that “a computer can never be held accountable; therefore a computer must never make a management decision.” This remains the fundamental guiding truth for the modern artificial intelligence era. Because an algorithm cannot face legal liability, possess moral conscience, or suffer consequences for a flawed decision, it can never serve as a genuine replacement for human authority. Ensuring that AI remains an augmenting tool rather than a substitute requires us to firmly anchor all critical workflows in human oversight. Whether in medical diagnoses, judicial sentencing, or high-stakes financial approvals, AI systems must only act as advisors that synthesise data and suggest outcomes, while high-agency humans retain ultimate responsibility, contextual judgement, and legal accountability for every decision made. If we build our systems while holding this phrase in mind, humans will not be replaced.