Will Larcombe, Co-founder and Director of Stellarmann , insists organisations can safely pioneer AI-enabled Insurtech investments, if they can identify (and draft in) the expertise needed to manage the associated risks and stakeholder expectations.
Few will have failed to notice the rise of artificial intelligence (AI) in Insurtech. Its emergence is driving rapid evolution in the way risk assessments, claims processes and fraud detection are conducted. It’s also weaving itself into general operations, helping to enhance customer facing functionality, and much more.
This is a technology wave which no one wants to miss. But, while the potential is huge, we need to approach it with a healthy amount of caution, and carefully assess how we deploy AI-enabled solutions. This is still a relatively new technology, with a significant number of “unknown unknowns” that are likely to present unexpected risk. So, before diving into transformation projects to deploy AI solutions, businesses should consider three fundamentals.
The first is capability. Given AI is an emerging technology, how can organisations source the experience and expertise needed to deliver an AI orientated digital transformation project?
The second is expectations. Is every solution going to deliver a return on investment within a reasonable timeline? According to KPMG, 62% of CEOs expect to see a return on their AI investments within five years, while 23% expect returns within three years. Is this realistic?
The third is risk. How can we safely move forward with solutions that may not be completely understood? Could these solutions compromise data privacy? Will we remain compliant within a regulatory landscape that is itself evolving?
Capability: attracting the necessary expertise
There’s no shortage of AI technology becoming available to organisations – and businesses will need to make a decision on whether to dip into this market or develop their own solution. The global AI insurance market is currently growing at more than 30% each year. The scope of these solutions is so enormous, the market is expected to be worth $80Bn by 2032.
Not every solution will be a silver bullet, however. Some will be more relevant than others. Each business will need to carefully assess whether solutions make the most sense for their specific operations, customer base and architectural framework. To navigate this space successfully and meet the expectations of senior leaders, organisations need to draft in people with the appropriate skillset.
The talent pool in this emerging area is still small and relatively undefined. Given the likely ‘unknown unknowns’, the roles and job titles required may as yet be undetermined. As such, finding people with these new skills and experience is the first step towards successful AI adoption. The next step is to attract and retain the best people.
This is already an incredibly competitive space – and organisations will be vying with each other to acquire this talent. Many of the roles will be quite niche, which means they’ll often be short term engagements, made on a project by project basis. Regardless of sourcing strategy, organisations still need to create an appealing and attractive project environment.
The businesses that are truly serious about AI and their future digital deployments will be the ones that succeed here. They will be the organisations that are creating a sense of excitement around what they are doing in this space. They will be talking about it publicly, highlighting the projects they are working on and building themselves a reputation for innovation. They will also be focused on creating the right conditions for success.
The best people will be looking for organisations that are committing resources to AI projects and, crucially, they will want to see buy-in from senior leadership.
Expectations: satisfying the stakeholders
Enthusiasm from senior leadership is essential. But, at the same time, so is a dose of realism. With the potential for AI seeming to be unlimited, and applications growing exponentially, it’s easy to get carried away with what can be achieved. Stakeholder expectations need to be managed though.
There may be a need for an initial education programme within organisations. Workshops with the senior management team and other key stakeholders could highlight success stories from across the industry and what will be possible going forward.
At the same time, it’s important to identify the solutions that make most sense for an organisation. There are literally hundreds of routes for businesses to go down with AI, so it’s essential that organisations prioritise.
There are several obvious factors to consider when choosing the best avenues to pursue. This includes the impact on customer experience, operational efficiency and the return on investment.
To build confidence in AI projects, project leaders should be looking to identify where quick wins are possible. Initially, these are likely to be those simple applications that remove mundane tasks and drive time saving efficiencies – be that in the automation of claims processing, the detection of fraud by identifying anomalies in claims applications, the analysis of data in order to forecast future claims trends, among many other potential applications.
By identifying the main areas of focus, in both the short and long term, organisations can start to build a better understanding of the investment needed. There will also be greater clarity across all stakeholders as to what is required to deliver projects successfully – what realistic timelines look like and what skillsets will be needed.
Managing risk: security, data and regulations
These initial conversations should also be framed around the organisation’s existing IT infrastructure and data management system. AI is dependent on access to high quality data, so significant thought needs to paid to how private and sensitive data will be processed – and what risk this creates.
A change advisory board or equivalent can support the AI project teams, advising what is needed from an information security and data protection perspective. They will also need to factor in the evolving regulatory landscape.
The UK is currently regulating AI within its existing regulatory bodies, but organisations operating in the US and Europe need to consider the implications of the National AI Strategy and the EU AI Act respectively.
Guidance on AI implementations from the European Insurance and Occupational Pensions Authority (EIOPA) may also need to be factored in. This covers everything from governance and ethical considerations to testing and cyber security.
EIOPA also encourages insurers to maintain high standards of data quality. This has to be a fundamental consideration as AI applications will only ever be as good as the underlying data.
To limit this risk, organisations need to bring in expertise to advise and deliver data cleansing solutions to ensure systems are working off the latest, most accurate data.
It would be foolish to take a ‘wait and see’ approach on AI applications. Yes, there are ‘unknown unknowns’. But if businesses are not demonstrating their innovative credentials, and working on exciting projects, they may find it impossible to attract the best talent in the future. By that stage it may already be too late to catch up with more forward thinking competitors.
In this sense, businesses must invest. They can still apply a degree of caution in their approach, however – by choosing the projects that are best for them, and being realistic about what can be achieved, in what timeframe – and keeping their stakeholders informed. By carefully considering their short and long term priorities, they will also know the skillsets they need to draft in in order to manage the risks, remain compliant and deliver a healthy return on investment.
Be the first to comment