The Socio-economic Impact of the Adoption of AI in Financial Services
Photo by Markus Spiske temporausch.com from Pexels
Artificial Intelligence (AI) and Robotic Process Automation (RPA) present new approaches to doing business, with potential to trigger innovation and the growth and productivity improvements needed to gain a competitive edge in today's modern, disruptive, digital economy.
A key change from previous industrial revolutions is that AI goes beyond automation of simple, mechanical, rote tasks to transformation of underlying business models, processes and the legacy technology stacks business increasingly relies on. Further, the scope of AI is increasing, with growing capability to carry out non-deterministic, knowledge-based tasks, as technology rapidly advances.
A number of viewpoints exist on merit for business, the economy and wider society, influenced by differences in understanding and application of the technology, and historical evidence of socio-economic impact of previous industrial revolutions.
Successfully harnessing AI and RPA requires fundamental transformation to legacy systems, models and processes, supported by long-term investment in rare, highly-skilled technology capability and strategic change leaders. Success requires also change to mindsets. This allows implementing organisations to adapt to new, agile ways of working across big data, to gain insight into customers, the market and the wider environment. Evidence shows that market leaders obtain real gains from improved process and risk management, and conversion of insight into effective design and delivery of high-quality, value-added services. An added benefit is organisational reputations that foster trust, aiding customer retention.
At the same time AI and RPA are set to disrupt the workforce with skill-biased technical change. Individuals possessing skills that both complement and support AI technology stand to gain. The agility that accompanies transformation to AI in the workplace augments task completion and supports flexible working practices. Further, because such roles require skills that are both difficult to acquire and require constant updating, such talent is highly sought and highly paid.
At the other end of the spectrum, significant advances in the technology increase potential to substitute digital capital for human labour. To remain relevant individuals must reskill, or risk redundancy or a move to lower-skill, low wage employment.
The resulting skill and income polarisation and potential for significant, long-term unemployment in a growth economy risks further disruption.
The study, for a dissertation for an Executive MBA at Edinburgh University's Business School, examined the wider socio-economic impact of adoption of AI/RPA, using the financial services sector as a case study, to determine:
- the scenario for changes in roles and skills within the workforce, and
- item potential wider socio-economic impact of adoption.
A mixed methods approach was used, employing, primarily, in-depth interviews with subject matter experts across financial services, technology provision and consulting and advisory roles. The study sought to identify challenges faced in transitioning to the new ways of doing business AI/RPA adoption requires, and to explore different perspectives on formulating strategy to manage the transformation process. Considering both opportunities and threats that the technology poses within the workplace and the wider environment in which the sector exists, the study examined also what other institutions and actions are necessary to prevent socio-economic disruption.
The findings highlight tradeoff between business profit and societal gain.
A number of recommendations were made for formulating strategy for adoption of AI/RPA that lessens the threat of disruption, working instead toward benefit of all stakeholders -- business, employees and the wider socio-economic environment.
Key Findings & Recommendations
AI/RPA present opportunities for business transformation that:
- leads to more effective, efficient operations
- fosters innovation that benefits customers, employees and other stakeholders.
AI/RPA pose threat of disruption
- at significant levels to employees whose skills are rendered redundant
- with repercussions beyond the business, that cascade to the wider socio-economic environment.
Current adoption of AI/RPA is typically hindered by:
- sandboxed experimentation that does not harness the full potential of the technology, in what is often seen to be vanity projects
- bolt-ons on incompatible and/or poorly designed technology stacks
- strategy that alienates employees by focusing predominantly on cost-cutting as a means to generate gains for the business and investors.
Successful adoption of AI/RPA however requires:
- alignment of strategic technology transformation with overall organisational strategy
- fundamental transformation of legacy systems, models and processes, along with mindsets
- effective investment and application to real business cases, to enable real return on investment
- long-term investment in human capital, technology and other relevant resources.
It is acknowledged that AI/RPA adoption will transform the workplace and workforce. While gains are clear, there is a risk these are eclipsed by negative consequences within and beyond the business. Responsibility for managing individual impact was seen to lie predominantly with the individual, supported by employers, government, civil society and education.
Strategies for managing the technology transformation start from the premise that business does not exist within a vacuum, nor is it sustainable outside wider society. Further, AI technologies rely on large data stores built using customer and other environmental data. Policy arguments are therefore being made at levels such as the EU to mandate fairer share of gains between business and wider society, through schemes such as digital taxation and the monetisation of data.
Collaboration or at least co-opetition between industry, government, education and R&D is seen to provide a strong base from which to advance AI (and other related) technology to the point where it enables productivity gains that relieve humans to seek new ways of working with technology or other employment, without sacrificing income.
Extending this further is a viewpoint that suggests decoupling leadership from a focus purely on economic gain. This requires moving to visionary thought leadership and strategy that focuses on the long-term and considers alternative, more inclusive metrics for measuring value, and that centres humans within the digital transformation process.