Probability & Partners: Balancing the risks and opportunities of AI
Probability & Partners: Balancing the risks and opportunities of AI
![Pim Poppe & Kevin Rojer (foto archief Probability & Partners).jpg Pim Poppe & Kevin Rojer (foto archief Probability & Partners).jpg](/l/en/library/download/urn:uuid:e9e987a1-8188-42b1-8dad-8d9d73fa962b/pim+poppe+%26+kevin+rojer+%28foto+archief+probability+%26+partners%29.jpg)
By Pim Poppe, Risk Management Professional, and Kevin Rojer, Service Line Lead AI in Risk Management, both at Probability & Partners
Artificial intelligence (AI) is transforming the financial sector by enhancing predictive analytics for risk assessment, uncovering market trends, and optimizing asset performance. These advancements lead to smarter decision-making in risk management and financial planning. However, many financial professionals have yet to fully embrace this AI revolution.
This hesitation is understandable, given the challenges and risks associated with AI integration. As AI continues to evolve, its integration into the financial sector brings both amazing opportunities and major challenges. To reap the benefits, it’s vital to incorporate AI seamlessly into existing frameworks while maintaining strict data security standards. By doing so, financial professionals can harness AI to drive innovation, boost efficiency, and improve the quality of their services.
In the financial sector, data security is critical because of the sensitive nature of the information involved. Financial institutions face the risk of financial losses, reputational damage, and non-compliance with stringent regulatory requirements. To mitigate these risks, institutions should prioritize secure data environments within their AI systems. This includes establishing robust security protocols, implementing strict access controls, using advanced encryption techniques, and conducting regular system audits to identify and address vulnerabilities. By securely managing both private and public data, institutions can mitigate risks, maintain client trust, and uphold their reputations, all while fully leveraging AI's potential.
We are not currently aware of any scientific research that specifically addresses how the capabilities of large language models (LLMs) are influencing innovation within the financial sector. It may still be premature to draw definitive conclusions in this area. However, earlier research from 2018[1] offers valuable insights regarding barriers to innovation that remain relevant today. We believe this study is also relevant for AI adoption by financial institutions. This study highlights thirty-one barriers, ultimately narrowing them down to six key obstacles that significantly impact the innovation process:
- Insufficient utilization of new ideas
- Inertia caused by local system architecture
- Unsupportive organizational structure
- Excessive risk aversion
- Absence of fundamental research and development
- Not-invented-here syndrome
In the financial sector, the insufficient utilization of new ideas is a noticeable barrier. Institutions often rely on established practices and hesitate to venture into uncharted territories. These can stifle innovation. This conservative approach, deeply rooted in the industry’s preference for stability and predictability, hinders the adoption of cutting-edge technologies such as AI. Overcoming this barrier requires a cultural shift towards valuing new ideas and encouraging creative thinking.
Inertia caused by local system architecture is another significant barrier, as entrenched legacy systems prevent the seamless adoption of new technologies. This rigidity can lead to inefficiencies and an inability to keep pace with rapid advancements in AI and other technologies. Transitioning to more flexible and modular systems can overcome this inertia, facilitating the seamless integration of new technologies and enabling institutions to stay ahead of the curve. By strategically updating legacy systems, financial institutions can create a more dynamic and responsive operational framework that supports continuous innovation.
An unsupportive organizational structure further complicates the innovation landscape. Without leadership that prioritizes a culture of innovation, employees are less likely to propose and develop new ideas. This lack of support can lead to a stagnant environment where the status quo is maintained, and innovation is discouraged. Creating a supportive structure that encourages experimentation and rewards innovative efforts is essential for progress. Balancing risk management with a willingness to take calculated risks is crucial to driving innovation forward.
Investing in fundamental research and development is vital for sustaining long-term innovation. Dedicated R&D efforts enable financial institutions to explore new technologies and methodologies, driving continuous improvement. Addressing the not-invented-here syndrome by encouraging collaboration with external partners and adopting best practices from other industries can also enhance innovation. This openness to external ideas can lead to the adoption of transformative technologies and strategies that drive progress and efficiency.
To truly capitalize on AI's transformative potential, financial institutions should prioritize experimentation and continuous innovation. Experimentation is key to discovering new AI applications that can enhance efficiency and effectiveness in the financial sector. By fostering a culture of innovation and encouraging pilot programs, institutions can conduct low-risk experiments. These small-scale controlled tests allow to explore AI’s potential without incurring significant risks.
Continuous innovation in AI systems requires a commitment to ongoing learning and adaptation. Financial institutions should focus on exploring new technologies and methodologies to maintain their competitive advantage. By adopting a proactive approach, institutions can refine AI applications and improve professional effectiveness. This involves regularly updating their knowledge on AI trends and advancements, ensuring the institution stays at the cutting edge of technological progress. Starting with small, manageable projects allows institutions to build confidence and expertise gradually, resulting in significant long-term benefits, despite the fast pace of AI development.
We are naturally observing what is happening around us, and by talking with our colleagues and clients, we can reflect on these previously identified barriers. These barriers are also obstructive in relation to the AI revolution. The current lack of freedom for independent thinking and categorical prohibitions against AI prevent us from fully exploring their potential. Adopting a risk-based approach to determine what should or shouldn’t be prohibited can unlock new opportunities for innovation. By always having a human in the loop to oversee AI systems, we can ensure responsible use while fostering an environment that encourages creative and independent thinking.
We have been experimenting and innovating with AI by developing our internal AI system using large language models (LLMs). Leveraging our existing IT infrastructure allows us to securely utilize our private data with these models. Our original approach of bringing the models to our data, rather than transferring data to the models, enhances security and operational integrity. This method ensures that sensitive information remains protected while still benefiting from AI capabilities. As a result, this strategy has led to significant improvements in our internal operations, boosting both efficiency and quality. Our commitment to continuous AI experimentation has enabled us to seamlessly integrate AI into our business processes while preserving the stability and security of our operations.
In conclusion, the importance of embracing AI as a transformative tool for financial professionals cannot be overstated. Creating a secure yet innovative environment is crucial to unlocking AI's full potential. Financial professionals should take bold steps to thoughtfully and securely integrate AI, beginning with small-scale projects and rapidly iterating for improvement. As AI evolves, the journey of learning and adaptation will be continuous. Looking ahead, the potential for AI to revolutionize the financial sector is immense. With strategic implementation, we can direct these changes effectively and sustainably, driving improvements in both efficiency and innovation.
[1] P. Das et al., Barriers to innovation within large financial services firms (European Journal of innovation management, 2018)