Everyone has an opinion about it, but not many organisations are actually doing anything. Therefore, […]
The impact of AI on the entire financial ecosystem
The new physics of financial services
Much has already been written about the role of Artificial Intelligence (AI) in financial services. But what are the strategic implications for the sector? How does AI change the operational models of financial institutions? And above all, how can institutions embrace AI to be better prepared for the future?
The World Economic Forum and Deloitte Global conducted extensive research to answer these questions. Seven global workshops were held over ten months, and more than 200 experts were interviewed. The result is described in the report The New Physics of Financial Services: Understanding how Artificial Intelligence is transforming the financial ecosystem. (Download)
So what exactly is Artificial Intelligence? Opinions are divided. Each definition differs to a greater or lesser extent. But according to the experts, AI is a series of technologies with a predictive power and a certain degree of autonomous learning. All in all, this makes us much better able to recognise patterns, anticipate future activities, create good rules, make better decisions and communicate with other people.
New focal points and new opportunities
The impact of AI on financial services is significant. Not only in a strategic sense, but also at the operational, regulatory and social levels. Artificial intelligence is changing the entire financial ecosystem. New focal points and new opportunities will arise.
Exactly how far-reaching the impact of AI is, immediately becomes clear when you recall the old ways of doing business in financial services. Large assets offered economies of scale. Physical locations and standardised products ensured cost-effective revenue growth. Direct access to markets and connection to investors discouraged competition. Switching providers was difficult, so customers usually stayed where they were. At the same time, process efficiency was a product of human effort and know-how.
AI is changing all these building blocks. Technology makes operations efficient enough. So the size of assets, although still important, will no longer be enough to build a successful business. Scale in data streams now determines the cost advantage. Neither is it any longer about standardisation, but about custom production and personalised interactions.
Exclusive relationships will no longer be a distinguishing factor. Providers have a good ability to create connections through digitisation. Customers now stay with an institution, not because it’s difficult to walk away, but because their benefits are better there than elsewhere. And process efficiency? This results from the interplay of human and artificial forces.
Eight ways AI is changing financial services
The shift in building blocks will have far-reaching consequences for the sector. Old business models will come under pressure. And companies built around the new components will exert that pressure. Once again, the entire financial ecosystem will change. And in no fewer than eight ways.
1. A new battlefield for customer loyalty
Historically, financial institutions relied on price, speed and access to attract customers. But online platforms make it easy to compare prices. Emerging technologies make direct delivery of products and services a basic expectation. And thanks to digital distribution, there is less need for intermediaries when doing business.
Old levers no longer work and new ones replace them. What is now effective is adapting offers to the customer’s specific needs and objectives. Involvement through continuous and integrated interactions that go beyond financial services. Such as offering forecasts to sellers or booking repairs for car damage. And curated ecosystems based on the data of consumers, business customers and other parties.
The new levers will ensure that financial institutions can compete strongly on value, retain customers, provide differentiated advice and offer one-stop solutions. All made possible by AI breaking the compromise between better service and cost.
But it can also sometimes mean competing with existing offers in different sectors. And companies may need to move forward before the natural price equilibrium in the platform economy becomes clear, or what margins one can expect to earn.
2. Self-driving financing
Financial advice, part of every product, is often generic and impersonal. It also depends strongly on the person you encounter in customer service. To improve financial results, it’s important to bring together information about products and customers. But that’s difficult, both inside and outside institutions.
That’s not the case with self-driving financing. Consumers contact an AI-based agent for advice and product modifications. This self-driving agent provides guidance on complex decisions such as buying a house, pension planning or corporate finance. At the same time, the agent automates routine transactions like billing and refinancing.
Artificial Intelligence enables self-driving financing in three ways. First, the technology can compare products and suppliers to achieve an optimal price and be suitable for the customer. It can also personalise the advice and products out there, more cheaply than human agents. And finally, a self-driving agent can manage daily finances fully behind the scenes. It avoids fees and monitors better deals without the customer having to intervene. But data is the driving force behind this scenario. Data from customers and third-party platforms, as well as from the financial institutions themselves.
So who will supply the self-driving agents? This is not yet known. Will it be established operators, newcomers or large technology companies? And how will the interests of consumers be aligned with those of product manufacturers and self-driving agents? In other words, what will be the accountability framework for decision-making, based on algorithms? Only the future will tell.
3. From cost centre to profit margin
Not only will AI change the front office, but the back office will also look radically different. For example, companies may convert their centres of excellence into services, while most other back office options are outsourced. Why? Because it’s difficult to excel at everything. And over time, competitors will probably mimic the processes that prove to be efficient.
At the same time, successful AI processes could improve so quickly that it becomes impossible for others to catch up. At such a time, a centre of excellence becomes both a defence mechanism and a sustainable source of income for the institution.
But AI also influences other parts of the back office. Intelligent technologies are emerging at a time when financial institutions want to modernise their activities, for example by using cloud-based architecture. Moving to the cloud makes it easier to have a plug-and-play with third-party services. It also makes it easier to turn internal centres of excellence into commercial offerings. And if the commercial offering is improved with AI, more data can be retrieved to learn from and keep improving continuously.
Early movers have the advantage in this scenario. In an open-source AI environment, it’s not so difficult to imitate algorithms. It’s especially difficult to collect large amounts of quality data. Existing Software-as-a-Service offers a blueprint for AI-based, outsourced services. But with this, established institutions still have to find out how centres of excellence can be built for an attractive offer. Another dilemma is how to protect the value of proprietary data, when that data needs to be shared with competitors for efficiency. The impact of data regulation on outsourced back-office processing may affect the globalisation of financial services. Finally, data security and cloud architecture concerns remain to be addressed.
4. Finding a balanced approach to talent
Talent is a sensitive subject in the context of Artificial Intelligence. The short-term focus of AI seems to be on doing the same things better. This could lead to the sector losing jobs faster than creating them. But if this happens without a long-term talent strategy, financial institutions could very well stumble when trying out larger AI opportunities.
The business models and competitive dynamics that AI generates lead to new opportunities for talent. And success in this new business reality will be determined by clearly different strategies around roles, culture and rewards. Those who don’t echo these trends may face a lack of capacity for innovation. A short-sighted focus on the short term. And strategies that reward the status quo. In short, a confrontation with a lack of investment in talent and technology.
Financial managers must thus ask themselves what else they need to do to manage their talent needs. What types of talent do companies need for new business models? And how can people keep pace with technological transformation?
5. Division of the market structure
AI reduces the costs for customers of searching and comparing. And that will ensure that financial service providers will find themselves increasingly at the extremes of the market. It increases the return for large-scale players. And it creates new opportunities for niche and agile innovators. With every institution fighting for a variety of data, more companies will join forces. With competitors and potential competitors. Medium-sized companies could thus become a thing of the past.
AI platforms will persuade customers to switch to cheaper providers when they realise that price drives many financial products such as loans and insurance. For customers with differing needs, optimisation algorithms find the niche products that suit them best. And AI makes it cheaper for companies to make products in responding to this demand.
All this can make competition difficult for mid-sized companies. They will have difficulty making the investments needed to stay in the game. Certainly when larger, established operators become AI service providers themselves.
So in this scenario, regulators must decide how to respond to the increased consolidation of assets. They can be put under pressure to reduce regulatory barriers to enable new entrants. And the open question is: to what extent will the consolidation of scale players become a cross-border problem where international firms expand aggressively in domestic markets?
6. Uncomfortable data partnerships
AI stands or falls with data. The more, the better. Partnerships can be a quick way to acquire the depth and breadth of data needed for more accurate models and more complex user cases. Access to end-users also increases the possibility of reliable data cycles.
But data partnerships will experience winners and losers. Some companies will be pushed to the periphery. Others will come to the fore as ecosystem hubs. Whatever happens, the AI opportunities that institutions focus on are likely to be short-term. While the risks of partnerships are long-term.
It’s a case of ‘Winner takes all’. Especially in platform and self-driving ecosystems. Winners gain excessive market power, allowing them to have providers bidding against each other. But they also gain more exposure to security and privacy risks that can break their partnerships. They may end up with declining strength if the data gap between major techies and incumbents continues to grow. And then there is partnership lock-in. Too much dependence on data flows from partnerships can continue relationships that ultimately turn out to be mainly bad.
Of course all these tensions are manageable. Other companies have already discovered how. Will financial institutions do the same? Time will tell.
7. Collective solutions for shared problems
AI can help with some of the challenges in the current financial system. For example, by addressing data asymmetries that impede fraud prevention, anti-money laundering or other processes aimed at a safer and more reliable financial system. A solution becomes even more important because a problem with the process at one institution can have a ripple effect on other institutions in the ecosystem.
These processes are rarely strategic. And moreover they are often generic across multiple product categories. Hence institutions may find it worthwhile to trade their own approach for the flexibility and efficiency of a mutual one. AI can then recognise patterns in the shared dataset and develop insights about threats that transcend institutional boundaries.
However in addition to leadership and investment, a collective solution requires several ways of keeping them in line with the interests of its stakeholders. There must also be agreement on how responsibility is shared for compliance errors and shortcomings. Finally, there is the question of whether cross-border solutions are possible given the diversity of financial and data regulation.
8. Regulatory and ethical dilemmas
Global data regulation is currently undergoing unprecedented change. Governments are moving to new rules to protect and empower citizens. And these rules affect the development of AI in several ways.
Firstly in cloud-based services. Regulation of financial institutions’ cloud usage varies globally, with stricter restrictions in Europe. In regions with more relaxed rules, technological players can develop new possibilities more easily.
The rules also affect AI in the use of personal data. New privacy and data protection limits the collection, transmission and storage of personal data. Data partnerships are becoming more difficult to manage as a result. Meanwhile consumers are gaining increasing control over how companies use their data.
Finally, there is influence through the rules on access to financial data. European regulations require that established institutions share customer financial data with other parties at the customer’s request. But data sharing is a one-way street. There is no need to share non-financial data with the financial institutions in return. This means that large technology companies can use financial data alongside a wealth of other personal data. And thus gain an advantage when developing new AI models.
Understanding and adapting to Artificial Intelligence is a journey of discovery. A journey that is subject to the headwinds and tailwinds of economic, social and political change. It’s also a journey that no company should embark on alone.
The future of financial services lies in its ability to take full advantage of new technologies. AI is a new technology that makes front and back office activities look completely different. It is causing major shifts in the structure and regulation of financial markets. And it offers society important new challenges. Nothing less than a joint effort will overcome these challenges. And unlock the benefits of AI for the best interests of business and society.