This is the year for AI in Insurance!
Now when have we heard that before? Why will it be different this time round?
Well… momentum was already building throughout last year with maturing AI Insurance offerings and then obviously ChatGPT launched in November achieving mass AI awareness and popularity amongst the public. Based on the feeling at the Insurtech Insights Europe conference last week, ChatGPT might have just been the spark to finally make 2023 ‘the year’. A larger number of the panels, conversations and sponsor booths had an AI slant to them, but the main difference was the energy and tone of business stakeholders on AI. ChatGPT’s impressive, user friendly demonstration of AI’s power provided a morale boost for sceptical industry executives and those in the front office roles fatigued by failed AI / intelligent automation initiatives.
Whilst not everyone is sold yet – there is optimism where there was once scepticism. That feels like one big obstacle navigated but now we ought to focus on delivering true, ongoing business value at scale. At the close of the conference, we had three main takeaways on how insurers can unlock real value when delivering AI-driven business transformation:
Many companies have previously attempted to drive AI powered transformations through top-down / “COE out” programs or have conducted isolated experiments within specific business units. The first of these typically receive good investment but fail to invest adequately in engaging the business units which often results in “tissue rejection” and transformation failure. The second has high business engagement but limited investment or necessary business change support (see Takeaway #2) leaving them stuck as ‘POC with promising results’.
The never-ending experimentation with model training POCs is starting to come to an end. Most industry executives recognise that AI is critical to transforming many parts of the business. The technology no longer resides in the sole custody of innovation and technology teams. Business teams now have a willingness, expectation even, that AI should be being leveraged to enhance their day-to-day activities.
Hopefully 2023 is the year digital transformation programmes become joined up initiatives to ensure all stakeholders have been bought-in effectively and there is sufficient investment of capital and resource to deliver at scale.
Many insurance companies have fallen into some of the classic traps we’ve seen with adopting InsurTech. They overestimate the ‘turnkey’ nature of the solutions and their companies’ ability to absorb change, attempting rapid AI transformation programs with little business change focus that leave their staff orphaned and disenchanted, ultimately leading to failure and perception that the ‘technology doesn’t fit’.
There were countless AI technology providers at the conference, and they seemed to be roughly characterised into two types: use case tailored or use case agnostic.
Use case tailored – The ‘Off the Shelf’ style solutions - Offering pre-built data models, workflows, integrations etc. All great accelerators IF you are willing to adopt them and adapt your business processes. Alternatively, you can configure and adapt the solution – either way there is work to be done.
Use case agnostic – The ‘Low/No Code’ style solutions – Offering a powerful and flexible AI technology paired with rapid training / configuration. Idea being you can rapidly develop the solutions to fit your business. The long pole in the tent here is normally not training the AI model to a ‘good enough’ accuracy (far too often the sole focus of POCs) but in successful operational integration.
There’s still a need to move in stages: understand what business improvements are strategic for you, identify where AI can enable those improvements, build a credible business case with a solid view on target state design and feasibility and build / mobilise a manageable roadmap of initiatives.
Put simply – delivering value from AI technology requires effective business transformation and a clear view on what the operating model needs to be.
For some, ChatGPT has signalled the beginning of a bright new world whilst others see it more as the start of a dystopian future. Insurers need to be mindful of how these technologies might bring various risks to businesses and society. How can you trust the accuracy and authenticity of what ChatGPT wrote, transparency on how the algorithms provided the information and potential bias / illegal use of protected characteristics in its responses? Greater governance and processes / frameworks (such as Deloitte’s Trustworthy AI) should be used when implementing AI solutions to ensure regulatory rules haven’t been breached and that these technologies are used in a transparent and ethical way.
These challenges aren’t necessarily new but now that business users are clamouring for AI solutions the time really has come to start delivering successfully and realising true value. As the industry grapples with expense ratio scrutiny, talent retention issues and an uncertain geopolitical and environmental future, the ability to successfully leverage the power of AI is critical. Harnessed and implemented well, it can deliver a critical competitive advantage as the new industry paradigm emerges.
We are really excited to see how insurance will transform in 2023 and to play a key role in this next wave of AI transformation.