Artificial intelligence has become one of the defining topics of our time. From boardrooms to universities, from government institutions to technology companies, discussions about AI have moved beyond technical innovation and into questions about governance, risk, business value, and even the future of humanity.
Speaking at the 3rd Credit Risk & Sustainability Conference in Nicosia, Giorgio Costantino, the Executive Director, Global Analytics & Consulting Services, CRIF Group, shared his perspective on how organisations can navigate this unprecedented technological transformation. Rather than focusing on artificial intelligence as a threat or a trend, the discussion centered on a more practical question: Can we manage AI effectively and transform it into meaningful business value?
A revolution unlike any before
As Constantino noted, throughout history, major technological revolutions have reshaped society. Industrialisation transformed manufacturing, electricity revolutionised productivity, and the internet connected the world in ways previously unimaginable.
Artificial intelligence represents the next great technological leap. However, unlike previous innovations, AI has sparked a different kind of debate. For perhaps the first time, society is openly discussing whether a technology should be slowed down, regulated more heavily, or even limited.
This debate has been highlighted by publications such as The Economist, which has explored concerns ranging from AI-driven geopolitical tensions to the possibility of long-term existential risks. Former researchers from leading AI firms have also raised questions about the pace and direction of AI development, the expert noted.
As Constantino pointed out, against this backdrop, the challenge facing businesses is clear: rather than asking whether AI should exist, organizations must determine how to use it responsibly and effectively.
Keeping expectations grounded
While enthusiasm surrounding AI remains exceptionally high, there is a growing recognition that expectations often exceed reality, the expert warned.
He noted that a study conducted by Boston Consulting Group found that approximately 80% of companies have introduced AI initiatives. However, among those organisations, roughly 80% reported no significant business impact from their investments. In many cases, implementation costs remain substantial while measurable value creation is still limited.
The encouraging finding is that a small group of organizations, often referred to as AI builders or AI leaders, are achieving meaningful results. Many of these organizations operate in financial services, fintech, and banking, demonstrating that AI delivers the greatest value when applied to clearly defined business challenges.
As Constantino told conference audienecs, the lesson is simple: adopting AI is not enough. Success depends on using AI to solve real problems and improve decision-making.
The global concentration of AI investment
One of the most important realities shaping the AI landscape today is the concentration of investment.
The development of foundational AI models requires enormous computational infrastructure and financial resources. Hundreds of billions of dollars are being invested annually, primarily in the United States and, to a lesser extent, China.
A handful of organisations, including OpenAI, Anthropic, and Google Gemini, currently dominate the market. Their scale and infrastructure create significant barriers to entry for competitors.
Europe has produced promising innovators, including companies such as Mistral, but the competitive gap remains substantial. As a result, most organisations should focus not on building foundational models themselves, but on leveraging existing AI technologies to create business value.
This is precisely where CRIF has concentrated its efforts, Constantino underlined.
Creating value through practical AI applications
As a global data and solutions company, CRIF's approach to artificial intelligence focuses on business outcomes rather than technological experimentation.
The company's strategy centers on three key areas where AI can generate tangible value for financial institutions and businesses.
Enhancing decision support through Agentic AI
The first area involves the use of Agentic AI to strengthen decision-making processes, Constantino noted.
One practical application is geopolitical risk management. Organizations operating across multiple markets increasingly face uncertainty arising from trade disputes, regulatory changes, political instability, and supply chain disruptions.
By combining AI with financial data and proprietary risk methodologies, CRIF can analyse where companies generate revenues, where they operate, and how specific geopolitical developments may affect their business activities.
For example:
- A company heavily dependent on the U.S. market may face exposure to changes in trade policy.
- Operations in Saudi Arabia may create sensitivity to regional developments.
- Investments in Chile may be affected by regulatory uncertainty.
- Supply chains connected to strategic regions such as Kazakhstan may introduce geopolitical spillover risks.
This approach enables institutions to move beyond macroeconomic analysis and assess geopolitical exposure at the individual company level.
Fighting fraud with AI
The second major area of investment is fraud prevention and deepfake detection, the expert went on to note.
As generative AI becomes increasingly accessible, fraudulent actors can manipulate documents, create synthetic identities, and generate highly convincing fake content.
The financial impact is substantial. Identity fraud and document manipulation continue to grow, creating significant losses across banking and financial services.
To address these challenges, CRIF has developed Smart Capture, an AI-powered fraud detection platform designed to identify document tampering and fraud attempts.
The solution examines multiple indicators, including:
- Metadata hidden within document files
- Image inconsistencies
- Data discrepancies
- Visual manipulation traces
- Copy-and-paste alterations
- Compression anomalies
- Noise pattern variations
The underlying philosophy is straightforward:
The most effective way to combat AI-enabled fraud is to fight AI with AI.
By leveraging advanced analytical capabilities, financial institutions can identify suspicious activity earlier, reduce losses, and improve operational efficiency.
CRIF is already working with banks and financial institutions across the United Kingdom and Germany to customize these capabilities for their specific risk environments.
The results have demonstrated the potential to prevent significant losses while accelerating verification and approval processes.
Trusted data remains the strongest guardrail
While AI continues to evolve rapidly, one principle remains constant, Constantino continued, adding: AI is only as good as the data that feeds it.
This belief sits at the centre of CRIF's strategy, the expert went on to note.
No matter how advanced a model becomes, unreliable information produces unreliable outcomes. Conversely, high-quality, trusted data significantly improves decision accuracy, consistency, and reliability.
For this reason, Constantino emphasises combining AI capabilities with trusted data sources, including bureau data, transactional information, and verified external datasets.
Partnerships with trusted organisations, such as Artemis in Cyprus, illustrate how data quality can strengthen risk management, credit assessment, and decision-making processes.
In many ways, trusted data serves as the ultimate safeguard against poor AI outcomes.
Building the Next Generation of Decision Intelligence
CRIF's vision extends beyond using AI as a standalone technology. The goal is to integrate AI directly into decision-making platforms that support business operations.
One example is Crescendo, CRIF's intelligent decisioning platform.
The platform combines:
- Trusted internal and external data sources
- Business rules and predictive models
- AI-assisted strategy design
- No-code workflow creation
- Portfolio simulation and testing
- Real-time reporting and performance monitoring
A key benefit of this approach is that business users remain in control. Rather than replacing human expertise, AI helps accelerate and simplify complex decision processes.
For example, institutions can upload policy rules, connect relevant data sources, and automate the creation of decision frameworks while preserving governance and oversight.
The result is faster implementation, improved consistency, and more scalable decision-making.
Focus on what you have
To illustrate the mindset required for navigating this technological revolution, Constantino referenced the inspiring story of Alex Zanardi.
The former Formula One driver lost both legs following a tragic accident but went on to become one of the world's most successful Paralympic athletes.
His philosophy was simple: Focus on what you can do with what you have, rather than what you do not have.
This principle applies equally to organisations confronting the AI revolution.
Rather than worrying about competing with global technology giants, businesses should focus on leveraging their existing strengths:
- Industry expertise
- Domain knowledge
- Trusted data assets
- Customer relationships
- Proven business processes
These capabilities remain powerful competitive advantages in an AI-driven world.
Three principles for managing AI
As organizations continue adapting to rapid technological change, three guiding principles emerge.
1. Fight AI with AI
Advanced AI tools must be used to detect fraud, identify manipulation, and strengthen security.
2. Build on Human Expertise
Industry knowledge and business experience remain essential guardrails that technology alone cannot replace.
3. Feed AI with Trusted Information
Reliable data is the foundation of effective decision-making and responsible AI deployment.
Looking ahead
The future of artificial intelligence remains uncertain. Predictions range from transformational economic growth to profound societal disruption.
What is certain, however, is that AI is already changing how organisations operate.
The challenge is not whether AI will shape the future, but how businesses choose to harness its capabilities.
For Constantino, the answer lies in combining advanced AI technologies with trusted data, strong governance, and deep industry expertise. By focusing on practical applications that generate measurable business value, organizations can move beyond the hype and use AI as a tool for better decisions, stronger resilience, and sustainable growth.
In the end, managing the AI revolution is not about slowing innovation. It is about ensuring that innovation is guided by trust, expertise, and a clear understanding of the value it can create.
