Customer service in the age of Artificial Intelligence
Artificial intelligence (AI) and related technologies have captured the imagination of financial institutions. But are there threats among the opportunities?
Since the financial crisis of 2008, there has been an acceleration in the digital transformation of financial institutions. During this transformation, data volumes have increased, new fintech competitors have emerged as have new business models, and cloud computing has opened up the possibility of much greater processing power. Digital transformation has also laid the foundation for many use cases of AI in financial services.
“AI gives banks the opportunity to differentiate themselves. If you are not different from the competition in the future, you will cease to exist.” Laurent Marochini, Head of innovation Societe Generale Luxembourg
The benefits of artificial intelligence (AI) fall into three broad categories:
Transformation enabler: AI can help banks to execute better, faster and at a lower cost. It also has the potential to improve risk management processes;
Improved customer relationships: via greater personalisation, and augmenting the role of human staff, AI can create better customer proposals;
New business models and revenues: AI has the potential to gather more data, enabling financial institutions to create new business models that increase revenues.
Many of the use-cases for AI are about doing more with less; about improving operational efficiency cost-effectively.
But AI also opens up a world in which financial institutions can differentiate themselves, improving customer relationship management by interacting more effectively with customers and gaining a better understanding of their behaviour. By understanding customers, new, better and, more tailored products and services can be created, increasing revenues for banks in a competitive environment.
The challenges of AI
Amid this promise, there is an elephant in the room – what impact will AI have on jobs and employment levels within financial services, and indeed on society as a whole? Conversations about AI and robotic process automation often centre on large-scale replacement of humans with technology. The spectre of social unrest, caused by widespread unemployment is of concern.
Rather than view AI as a threat to human staff, financial institutions can deploy AI to aid staff and augment human capabilities. AI tools can help employees to navigate through the complexity and volume of data. A financial institution will always need human expertise because every client’s financial situation is unique.
In compliance, financial institutions will not be able to rely solely on automation. Output data from AI systems is probabilistic; there is never a ‘yes’ or ‘no’ answer so bringing a human to review AI output is required, particularly when it comes to regulatory reporting.
Another threat raised by AI is the growth of cyber-attacks. AI is not the sole preserve of financial institutions – fraudsters and cyber attackers also have access to open source AI algorithms. Cyber security will be an important element in the development of AI in financial services.
Financial institutions should also remember that AI cannot do everything; it is not a magic wand. We should not expect too much of AI – there are some cases where it will be very powerful but other areas where its value is yet to be demonstrated.
The potential benefits of AI for operational efficiency and business models have sparked significant investments in technology and a race for talent, particularly data scientists and analysts. However, to unlock the full potential of AI and give a significant competitive advantage, the technology must be combined with other digital levers, such as open banking, digitisation, social media and potentially blockchain.
The ‘big tech’ companies, which have mastered these different components from their inception, have had data at the centre of their business models. This sector presents a new competitive threat to banks and such organisations are coming into the financial services market very quickly.
Banks that do not adopt AI at scale might become laggards, operating at higher cost and unable to provide customers with key services and high-quality user experience.
Towards responsible and ethical use of data
As an increasing number of AI systems are developed, financial institutions must ensure that they implement the correct governance of algorithms, deploying a good enterprise-wide data framework and ensuring data, along with models and algorithms, and ensure that related data is properly generated, stored and can be audited.
If we let AI proliferate across the financial institution without such governance, it will be difficult to answer regulators’ and clients’ questions about the use of data.
“AI is a key transformation enabler for banks, enabling better, faster and cheaper execution but also opening new business opportunities once at scale but require to be paired with a responsible approach” Julien Molez, Group Transformation Data & AI Director, Societe Generale