Crisis breeds innovation. This might be a controversial statement and not obvious but think back to post the financial crisis – Uber, Whatsapp, Cloudera, Slack, and Square (to name but a few) were all founded in 2009 alone. We could easily go on, specifically when talking about challenger banks and payment providers.
The current pandemic has deconstructed many of our societal patterns, brick by brick, and challenged existing, traditional systems. When one system is close to collapse, another has a chance to rise, so AI and FinTech have been dramatically rising in power, as conventional solutions failed to meet many people’s expectations. Research conducted by deVere Group revealed that the use of FinTech apps in Europe increased by 72% during the pandemic, and so did the general interest in FinTech solutions. While FinTech is not immune to people’s fears and market volatility with PayPal’s drops in e-commerce activities and Mastercard’s and Visa’s decrease in predicted revenue, COVID-19 may be a golden opportunity for FinTech companies to truly win over customers’ trust and money. Concerns for public health certainly accelerate cash’s fall into obscurity and digital payments’ rise to global prominence. In 2019, the worldwide adoption of the rate of FinTech services was already at 56%, but the virus brutally exposed the weaknesses of traditional financial systems, highlighting the need for contactless transactions. Nevertheless, FinTech’s impact is greater than just the way we shop. Credit management risk, designed to mitigate risk and create a secure banking system, is on everybody’s radar now that uncertainty is the global daily bread. Could technology accurately predict which customers are not trustworthy? Experts claim that with real-time scoring and monitoring limits, banks could efficiently minimize their losses and flawlessly assess customers’ financial capabilities. Now, many people wonder if such deep insight into our private lives and behavioral patterns does not violate our right to privacy, and such concerns are raised for a reason, given the impressive accuracy of AI’s judgments. Interestingly, according to Statista’s survey, in 2019, 75% of Artificial Intelligence’s use was security-related, suggesting the vast potential of technology to both compromises and protect our privacy.
How will AI and FinTech shape our global future, from the way we shop to the way we are assessed by organizations? FinTech & Financial Services Consultant – Paweł Sobotkowski asked the Thought Leader and Author Michael Berns, to share his insight into the current situation of the FinTech market, potential future scenarios, and the technology’s strengths and weaknesses. Michael is a Director for AI & FinTech at PwC with 18 years of international experience and a passion for innovation in the global startup ecosystem. His PwC study „AI in Financial Services” has just been released earlier this week. And „The AI Book”, which he co-authored, was released last week. Thought Leader Michael will share his unique perspective and guide us through the complexities of AI and the FinTech world.
Let us embark on the journey into the future of finance!
Financial crisis may be Fintech’s door to global dominance
Crashing the economy, rising unemployment, and shifting financial models should work to Fintech’s advantage according to our expert, Michael Berns. In the post-pandemic reality, the market share for Fintech is expected to grow, including “peer-to-peer lending and crowdlending”. Michael believes that COVID-19 is driving demand for loans, and in turn, it creates an opportunity for quickly gaining market share.
Key takeaway:
The current crisis shifts the market for loans towards digital and scalable players, where we see some of the more digitally enabled banks but also more and more FinTechs.
The fine line between our right to privacy and AI’s behavioral analysis of social media accounts
Artificial Intelligence’s use of our social media behavioral analysis in Credit Risk Management has sparked up a heated debate on the nature of data available on our social media, and the line between humanity’s desire for innovation, and an individual’s right to privacy. In Michael’s view, “as with all AI / GDPR related use cases the key is to make the value of personal data transparent to the owners and let them decide”. He points out that the Western attitude towards privacy issues is different from the one adopted in Asia, where attractive payment rates and convenience outweigh the importance of personal data protection. Michael predicts that the Asian privacy model will quickly gain prominence in Europe, as pandemics have reframed the way we think about data privacy, especially healthcare data.
Key takeaway:
We will see a cultural shift in Europe to fully accept giving up some of your privacy in order for other benefits as those benefits have yet to be made transparent to us. Once these benefits are clearly defined and the trade-off is transparent I have no doubt there will be more and more opt-in offerings even in Europe.
Nothing can guarantee absolute freedom from social and racial bias
According to Michael, no one can guarantee that AI will be free of bias, but companies can work on fair AI integration into society. Our expert himself has been developing special Responsible AI frameworks at PwC and has contributed to a research report on unconscious bias addressing the potential of the new technology.
Key takeaway:
Most companies are looking to achieve the benefits of AI, so they will typically first need to get to critical mass in terms of clean data, and fine-tune for bias, etc later.
Even algorithms can make mistakes during the credit risk assessment process
Michael believes that even the most advanced algorithms can be prone to overfitting, and the best results are actually produced by the collaboration between a man and a machine.
Key takeaway:
The overall quality of the model is still determined by the old gentlemen’s rule of „garbage in garbage out”. In other words: even the current process relying on clean and comprehensive data can be prone to overfitting. The same can happen with a Machine Learning algorithm.
Artificial Intelligence as an early warning system for the crisis
Having run 100,000 Monte Carlo Simulations to calculate “a more accurate Value at Risk”, Michael is convinced that “the correlation between the factors historically observed was completely different from the actual behavior of markets during the crisis”, so AI algorithms could make inaccurate predictions solely because of the changing correlations between the factors. In turn, it would be more beneficial to use AI as an early warning system for the crisis, rather than just a tool to accurately predict market risk.
Key takeaway:
If we try to learn from the lessons of the last crisis then ideally we are using AI to create a holistic early warning system and spot hidden correlations rather than trying to accurately predict market risk be it via Expected Shortfall, VaR, or any other metric.
Artificial Intelligence can reduce the rate of human error but not without the human element
Our expert recalls that he has replaced “many legacies (rule-based) systems in Top 10 Global financial institutions with AI solutions”, as “the Machine Learning based solution was significantly outperforming the existing solutions, from factor 2 to factor 40.” He goes on to add that innovative technology can improve financial systems, but its success rate truly depends on people’s governance and staff’s competencies.
Key takeaway:
AI often can protect banks from human error much better than existing systems, but it is really the combination of strong governance, solutions, and skilled people that will reduce the error rate dramatically.
Autonomous Response is AI’s superpower
Based on his experience at Level39, Europe’s biggest FinTech, Michael lists Autonomous Response as one of AI’s most innovative uses. It allows the AI tool to extremely quickly decide about the best response to the event and act accordingly.
Key takeaway:
The most innovative methods include Autonomous Response, similarly to self-driving cars which react in extremely short time spans, the AI solution decides and then acts out the most appropriate response.
When traditional solutions cease to work, innovation is suddenly invited to the front rows. This is precisely why our expert, Michael Berns, sees COVID-19, which is “driving the demand for loans”, as an opportunity for FinTech companies to quickly gain popularity and market share. In his view, the pandemic will potentially accelerate the adoption of FinTech, for example, credit risk management, which makes use of available social media information to analyze customers’ behavioral patterns. Michael genuinely believes in AI’s potential, but he is aware of its weaknesses and believes that the key to success is the mix of “strong governance, solutions, and skilled people” rather than just technology itself. Indeed, the future of AI and FinTech looks bright, but possibilities need to be supported by leaders and regulations to become more than mere visions. If people and law are with AI and FinTech, who can be against them? Only time will tell.
