The rules of the financial industry are being rewritten thanks to AI and generative AI. In their conversation for The Bridge, Artefact’s Joffrey Martinez, Managing Partner & Global Lead for Financial Services, and Alexis Baufine-Ducrocq, Partner & Lead Financial Services, discuss the race to redefine financial performance with AI.
Joffrey Martinez oversees data and AI transformation programs for major banks, insurers, and financial institutions worldwide. Before joining Artefact, he worked in private equity and consulting, including at BearingPoint and IBM Consulting France. He holds a masters from Kedge Business School and Certification Black Belt Lean Six Sigma, Executive Education from École Polytechnique.
Alexis Baufine-Ducrocq helps leading banks and insurers deploy AI and GenAI solutions. He has over 14 years of consulting experience across Europe and Asia and previously held roles in strategy and technology. Alexis studied engineering and telecommunications at CentraleSupélec and is a graduate of ESCP Europe.
Artificial intelligence has long been part of the financial world, but the acceleration of generative AI and advanced analytics is now redefining how institutions think about performance, productivity, and value creation. Across banking, insurance, and asset management, AI is becoming not just a tool for operational efficiency but an engine of strategic transformation.
Today, competitiveness in finance is measured not only by margins or risk exposure but by how effectively institutions can capture and use their data. The race is on to deliver faster decisions, better customer experiences, and more resilient operations, all built on the intelligent orchestration of information.
Speed, accuracy, and quality have become inseparable. In an environment where a bank can approve a loan in hours rather than days or an insurer can reimburse a claim within minutes, customer expectations have permanently shifted. Leadership in finance will belong to those who can combine agility with trust: a combination made possible by AI.
Two paths to AI maturity
While every financial institution is engaged in AI transformation, the level of maturity varies widely across the sector. Retail banking and insurance were among the first to industrialize AI, largely because their business models depend on scale: millions of customers, billions of data points, and constant pressure on margins. These conditions make the return on AI investment both measurable and immediate.
Retail banks now use machine learning to pre-qualify loan applicants, personalize offers, and detect anomalies in customer transactions. Insurers use similar techniques to predict policy renewals, detect fraud, and optimize underwriting. In both sectors, AI directly supports profitability by reducing operating costs and increasing cross-sell potential.
By contrast, investment banking, private banking, and asset management entered the AI era through specialist domains: quantitative trading, risk modeling, and fraud detection. These uses were deep but narrow, confined to experts such as quants and actuaries. What is changing today is the diffusion of AI across the value chain. Customer relations, credit granting, and risk management are now infused with predictive models and automated reasoning. What once lived in isolated centers of excellence is becoming embedded in everyday decision-making.
As Joffrey puts it, “All the players have jumped into the race for AI, because there is a real competitive challenge in unlocking this potential.” The diffusion of AI marks the transition from experimentation to integration, from isolated projects to an industry-wide capability underpinning competitiveness.
Building the data foundations
Among financial subsectors, insurance offers perhaps the clearest picture of how deeply data defines transformation. Actuaries, the original data scientists of the industry, have always used statistical modeling to price risk. That culture of quantification provides insurers with a natural advantage as they modernize their infrastructure.
In recent years, the focus has shifted toward creating end-to-end data ecosystems. Health, property, and casualty insurers are building platforms capable of ingesting customer data, claims data, behavioral data, and external sources in real time. The result is a unified foundation for prediction and personalization.
Most insurers now operate on hybrid, multi-cloud architectures. Highly sensitive datasets, such as medical records, remain on-premises, while less critical data migrates to public clouds for scalability and advanced analytics. Increasingly, a third layer, sovereign clouds, is emerging to reconcile innovation with regulatory requirements. This tiered approach allows firms to balance compliance, agility, and cost.
Equally significant is the rise of robust data governance. Ownership of data is moving closer to business functions, giving underwriters, claims managers, and marketers direct accountability for data quality and usage. Strong governance turns data from a technical asset into an operational one: clean, contextualized, and usable for AI. With strong governance, AI moves from expert tools to self-service systems that accelerate business impact.
Customer Data Platforms (CDPs) are becoming central to this strategy. By linking information across touchpoints, CDPs enable insurers to detect early signs of churn, tailor offers, and respond dynamically to customer behavior. In a sector known for high attrition, where, as Alexis puts it, “the game is to fill the bathtub faster than it empties”, data integration is now synonymous with customer retention.
With such foundations in place, AI can move from expert-driven initiatives to self-service applications accessible to every business unit. Data thus becomes the bridge between technological potential and commercial impact.
Global contrasts in AI adoption
When comparing AI maturity worldwide, clear patterns emerge. The United States leads by a significant margin, accounting for more than a third of global AI revenue in financial services. American institutions such as JPMorgan have embedded AI into nearly every process, from trading and credit management to client advisory, demonstrating what full integration looks like at scale.
Asia follows closely, driven by China’s extraordinary capacity for innovation. Ping An, for instance, has evolved from an insurer into a diversified technology company, filing thousands of AI patents and industrializing hundreds of use cases across healthcare, lending, and wealth management.
Europe ranks alongside Asia in the scale of its initiatives but distinguishes itself through its emphasis on ethics and regulation. Banks such as UBS, BNP Paribas, and Santander are investing heavily in responsible AI frameworks while simultaneously deploying advanced models in fraud prevention and customer service.
Other regions are advancing rapidly in distinctive ways. Latin America, led by Brazil’s Itaú and Nubank, is using AI to combat fraud and manage digital-only banking ecosystems. The Middle East is benefiting from state-driven AI investment strategies that accelerate innovation across sectors, while African institutions are using AI to extend financial inclusion to underserved populations.
Despite these differences, one consistent factor unites the frontrunners: the recognition of AI as strategic infrastructure. Nations and institutions that treat AI not as a cost lever but as a growth platform are moving faster, more confidently, and with greater coherence.
Three vectors of value creation
Across the financial sector, AI is creating value through three interconnected levers: cost optimization, revenue growth, and risk protection.
Cost optimization remains the most visible. Automation is reducing the “cost to serve” by streamlining document processing, underwriting, and customer servicing. In retail banking, AI-driven credit analysis has shortened loan-approval times by days, reducing operational overhead while enhancing customer satisfaction. In insurance, end-to-end automation can now process up to 30 percent of property and health claims within minutes. Optical character recognition and generative models read invoices, verify coverage, and trigger instant payments: a tangible example of efficiency translating into experience.
The same systems elevate human contribution. By handling routine cases automatically, AI frees experts to focus on complex or sensitive situations where judgment and empathy matter most. Automation, in this sense, becomes a way to re-humanize work rather than eliminate it.
Revenue growth represents the next frontier. Predictive analytics enables banks and insurers to anticipate customer needs – a new job, a home purchase, a life change – and propose relevant products at precisely the right moment. Generative AI extends this advantage in advisory contexts: in private banking, it can condense complex market analyses into concise, actionable insights, reducing the delay between expertise and client decision.
Such capabilities transform service models from reactive to proactive. Relationships become continuous, contextual, and personalized, representing a significant differentiator in markets where product offerings are otherwise commoditized. AI sets a new standard for customer relations: available, accessible, personalized, and responsive at any time.
The third vector, protection, is perhaps the most critical for trust. Machine-learning models now monitor millions of transactions in real time, detecting anomalies that signal potential fraud or money-laundering activity. AI’s ability to handle unbalanced data, where genuine transactions vastly outnumber fraudulent ones, has revolutionized risk detection.
In compliance, AI dramatically reduces false positives. Instead of armies of analysts reviewing transactions manually, AI filters alerts to focus human attention on genuinely suspicious activity, reducing them in some institutions by a factor of 100 with no loss of accuracy. The result? Lower costs, faster response times, and greater regulator and customer confidence.
Beyond ROI: the cultural return
Quantifying AI’s impact is relatively straightforward. Fraud models can save tens of millions of euros; process automation can deliver 20 to 30 percent efficiency gains. Yet the deeper value lies in how AI reshapes the way organizations think and work.
Treating AI purely as an automation tool misses its transformative potential. It is not just about saving minutes on routine tasks; it is about redefining roles, management practices, and even the concept of expertise. In tomorrow’s insurance company, for example, managers will supervise AI agents handling most standard claims, dedicating human time to exceptional cases that require empathy and judgment.
There is also the quieter productivity of everyday tools, which Alexis describes as diffuse productivity. Assistive technologies like Copilot may save only a few minutes per user, but their cumulative effect harmonizes quality across teams. When employees share discoveries and best practices, they accelerate adoption and build confidence in more complex AI applications.
In this way, cultural readiness becomes both a prerequisite and a product of AI success. Organizations that foster curiosity and collaboration capture far greater long-term returns than those that focus solely on immediate financial metrics.
Managing the new dimensions of risk
No sector is more experienced in risk management than finance. Yet AI introduces novel forms of risk that require adapted controls, from data privacy and model bias to cyber-resilience and explainability.
Many institutions are addressing these challenges through the “three lines of defense” model. One team develops the algorithm, another validates its integrity, and a third, often including regulators, reviews its performance. This separation of duties ensures accountability and transparency across the AI lifecycle.
Crucially, the industry’s ingrained culture of governance provides a strong foundation for responsible AI. Banks and insurers already operate under rigorous standards of confidentiality and compliance; extending these principles to algorithmic systems is a natural progression. As Joffrey notes, “The greatest risk is the risk of not doing anything.” Inaction, in this context, would mean ceding competitive ground to more agile rivals.
The human transformation
The most profound change, however, is human. History shows that every wave of automation shifts professional focus rather than erasing it. When the mechanical calculator appeared in the late nineteenth century, clerks were not replaced; they became analysts. The same logic applies today.
AI will automate calculations, documentation, and monitoring, but the need for human oversight, interpretation, and empathy will only grow. The professionals of tomorrow will be supervisors of intelligent agents, orchestrators of processes, and designers of transformation. Their value will lie not in routine execution but in insight and judgment.
The challenge is not to make humans obsolete but to make them indispensable in a new way.
Toward a more intelligent financial ecosystem
AI has already reshaped how financial institutions manage operations, engage with clients, and control risk. The next phase will see specialized AI agents operating end-to-end alongside human experts, embedding intelligence into every process.
The institutions that will lead this transformation are those capable of combining innovation with the discipline that has always defined finance. They will see AI not merely as technology but as a means to renew the fundamental contract of the industry: to serve customers better, manage risk responsibly, and create value with intelligence and integrity.
In short, the future of financial services will not be decided by who deploys the most models, but by who uses them to build the most trusted, adaptive, and human-centered institutions.
Watch the original interview in French:
