AI entered the business, public and regulatory consciousness in 2023. No longer a covert, hidden component of technologies, but an overt and active component.
This growing awareness of the power of ‘overt’ AI has created a critical time during which financial services leaders must devise their approach to AI adoption.
The opportunity for AI growth more broadly in financial services might be new, but transformation, change and new opportunities created by disruptive technologies are not. This combination of emergent change combined with the learnings from previous paradigm shifts creates a clear mandate for action.
From our work with global financial services organisations, there have been some key learnings where businesses should begin when it comes to their AI strategies.
AI, like digital technologies before it, is set to create societal shifts. Regulation is therefore emerging in this space that will set the tone for how the financial services industry must adapt and respond.
The EU’s AI act is set to be the first past the post in terms of jurisdictional regulation with provisions becoming binding as soon as 2026.
The act, which includes rules from the use of chatbots to areas such as facial recognition, is likely to become the benchmark as other nations position their own governing standards surrounding where and how AI is applied.
As with GDPR, businesses who don’t have a well-defined approach in place by the time the act is enforced could face far-reaching repercussions. Organisations could be looking at punitive fines of up to 7% of their global revenue – and if you’re a bank that’s operating with multi billion revenues it’s easy to see the impact this could have on your business.
The two year transition period ahead of the EU AI act becoming law gives businesses the opportunity to get their house in order – and with AI features becoming ingrained into organisations' existing application portfolio, doing nothing isn’t an option.
While the EU is largely taking a prescriptive approach to the use of AI as it sets out how to protect society, in the UK the FCA is taking a more principle-led approach as it starts to set out legislation supporting innovation opportunities.
As well as considering the risks, the FCA is also looking at the opportunities and how financial businesses operating in the UK can maintain a competitive edge, while still ensuring the stability of the financial market.
The digital revolution led to financial services organisations creating disparate solutions with a lack of coherence in approach, not always building the right foundational enablers. Many financial services organisations have spent the last few years unpicking these, whether seeking to consolidate cloud infrastructure, or build fit-for-purpose mechanisms to innovate at pace with right sized governance.
In order to learn from this, an AI strategy that clearly articulates: the AI opportunity, investments that must be made, the direction for AI at your organisation and the key business value that is to be created - is integral.
The AI landscape is subject to rapid change. From our work with highly-regulated financial services organisations, we’ve seen success for businesses with a strategy that is not fixed to a point in time. Not only outlining what their position is now, but also how they can adapt over the coming years as the technology evolves and becomes even more disruptive and ingrained.
With AI technology changing rapidly, (even the change and advancement in ChatGPT alone), being overly prescriptive means you could be out of date in 12 months’ time. And the same goes for your AI strategy.
Clear direction and where to start will allow you to utilise the technology to underpin good decision making as well as identifying and mitigating the associated risks. It will also set the framework for establishing appropriate controls and capabilities internally to manage that risk, as well as giving you the confidence to both understand and take advantage of the opportunities.
Taking this into account, we advocate businesses taking a principle-led approach to both protect from risks and to ensure they can pivot and thrive in what will continue to be a rapidly changing landscape for some time.
AI, just like people, is great at problem solving. However, just as your hiring strategy needs to focus on the specifics of the roles and value you want to create, so must your approach to AI. Define your value levers and consider what you want AI to deliver for your organisation.
Perhaps it can give you productivity gains by automating some tasks, such as document processing, that will then allow your people to be more productive and focus on more interesting and valuable work. Equally it could open up new revenue streams by launching new services that you wouldn’t typically be able to do, or do them in a fraction of the time. It could also create a better customer experience by allowing you to do things that are more personalised and relevant to the end user.
You may also want to improve your organisation’s regulatory footprint by having AI look at risk at a more macro level than humans can.
Once you have decided on which value levers to focus on first, linked to key business objectives, you should identify the opportunities these are linked to, providing a true understanding of how you can use AI to transform your business.
Businesses should start by defining what AI is and is not. The OECD has a sensible definition and is a good place to start. From here, you can categorise opportunities into AI and not AI. This will help govern the methods and controls you put in place through planning, delivery and operations for AI powered solutions. This classification is important due to the provisions set out in emerging regulation such as in the EU AI Act.
A transformative technology such as AI isn’t immune from the need to do discovery. Aligning with your overall business objectives, and the specific challenges and goals of business units, teams and individuals. As a change agent, AI solutions also need embedding in your business, with corresponding changes to processes, ways of working and individual capabilities. Discovery is key to understanding all of this. Find pain points and ask what conceptually you can do to solve these problems.
Your AI strategy should be aligned to your key business objectives and key measures of success and this should guide prioritisation and where to focus investment. It should also align to the individual team business unit goals you have discovered.
It's also critical to understand where not to apply AI where simpler, less onerous solutions to problems exist
AI does not exist in a vacuum and it’s imperative that you bring other aspects of your business into the process early as part of establishing AI readiness and as a component of the development of new AI solutions.
Compliance is a critical factor when working in a highly regulated sector. Consider what you already have in place when it comes to governance and whether new plans fall within regulatory boundaries. This will also help you to set a baseline of where you are as a business and your readiness to adopt AI.
If you don’t have well-structured, well governed mechanisms to build and operate AI solutions you may find yourself liable to regulatory non compliance, financial hits and reputational damage.
Data quality is a shared issue amongst large organisations, but the impact of poor data quality can be amplified by AI, as models will be making decisions based on poor foundations. Take a vertically integrated approach, where the data quality attributes needed by your AI models are investigated, improved and monitored alongside AI development.
Factor in requirements around explainability in terms of the specifics of the problem you are trying to solve with AI and fit for purpose solutions. Generative AI, while powerful, is non-deterministic and its decision making can be hard to explain. There are approaches to increase explainability, so ensure it is a key component of any decisions around fit for purpose solutions and how they will be monitored.
AI is going to change how people work, but rather than let fear drive the conversation, start consultation early about how AI is going to impact individuals and teams. While ways of working may change as your business embraces this new technology, you can reassure them that your approach is more co-pilot than autopilot. This approach also creates AI opportunities, as employees are enabled and empowered to be AI assisted
Don't go for anything that’s high risk or in a critical part of the business when there are likely to be plenty of low hanging fruit that can be used as you pave the first path to production.
When you’ve decided which opportunities should be prioritised, you can take a hypothesis driven approach to validate the performance, impact and benefits to a specific problem you’re trying to solve.
As your capabilities are extended and your governance and control functions mature to support broader adoption of AI, you can start moving up the risk spectrum and you’ll witness a compounding return on the investment you made in your early use cases taken to production.
With AI built into so many applications that are used every day, businesses must act now to avoid being left behind in a tidal wave of disruptive technology. An AI strategy should be anchored on where the opportunities are, what capabilities are required to win and how to mitigate associated risks.
Getting expertise can help when it comes to manoeuvring what is a constantly evolving space. If you’d like more information on how an AI strategy can transform your business contact us: https://www.mesh-ai.com/contact
.png)
.png)
