Become a TechRadar Insider
The insurance sector has tended to be more cautious in technology adoption.
Understandably careful about the hugely sensitive data they hold on customers, insurers have traditionally taken a ‘you go first’ stance, waiting in line between early adopters in other segments.
But today, they confront an existential mandate to move to the Cloud and AI tools.
Fast.
CTO, Sapiens.
AI is going to be utterly transformative for insurance. It can automate rote tasks like claims processing that chew up worker time and create errors. Agentic AI creates new workflows in record time so pilots can go into production in days or weeks rather than months or years.
AI can assess the massive risk calculations involved in providing cover for innovative technologies such as humanoid robots or industry sectors like renewable energy.
Sceptics may suggest AI judgment is not perfect, but such is the velocity of its progress that it’s tempting to counter ‘no, but it will be in three months’.
The inconvenient truth about AI
The essential, undeniable, and even for some naysayers, inconvenient truth is that AI is letting insurers move faster than ever before and advance by orders of magnitude, not with the usual ‘steady as she goes’ iterative progress.
The ongoing breakneck pace of AI and the latest developments, such as a vast crop of enterprise plugins and Claude Code for software development, have changed the traditional dynamic of IT adoption. Insurance institutions that ‘get it’ realise that if they don’t move to embrace AI, they will be left behind and perhaps be confined to the dustbin of history. We are witnessing a warp-speed version of what happened with the dotcom coming of age. Things are happening very fast and will never again appear so low.
For insurers that get AI right, there are glittering prizes. Some of them are just IT automation writ large: McKinsey’s Insurance Productivity 2030 research discovered that up to 40% of a commercial lines underwriter’s time is consumed by rekeying data, conducting manual analyses and other administrative tasks. The more interesting prospect is how AI recalibrates the rules of insurance risk calculation, and nobody wants to be late to that party.
But making a flight to AI, like any strategic shift, requires planning and change to IT management. As insurers plan how to embrace AI, they must address several critical factors by answering four key questions…
First, have you moved to Cloud?
The insurance industry has long relied on highly customised legacy systems. But if it wants to make the journey to AI, then getting off that old horse and adopting the Cloud is the most appropriate new vehicle. Cloud storage and the SaaS paradigm of regularly updated configurable technology are a perfect match for AI’s rapid progress. Scale and agility are critical to enabling insurers to use AI to respond dynamically to volatile geopolitical and climate events.
Equally, though, the discipline of putting the right foundations in place to move the Cloud, including a master data management strategy that emphasises the importance of data quality and a prioritisation of integration between core applications to ensure a single view of customer information, will enable AI tools to perform with greater effect.
Second, do you understand the outcomes you want to achieve?
It is easy to become caught up in the industry, media and marketing hype around the potential of the technology. But unless you have a clear and documented blueprint for the outcomes you want to achieve, your AI big bet will not deliver value to your organisation.
Outcomes may be related, for example, to a greater pace of innovation in new products for commercial P&C customers, more accurate reinsurance programs, better handling of volume transactions, lowering risk thresholds or liberating people to do more creative work.
Third, do you recognise that core systems are the oil that powers machine learning?
The shiny new machines are great, but unless AI is connected to a core system, it is nothing but an adviser that doesn’t have the data to power decision-making.
And perhaps more concerning, if you have not implemented a data strategy to ensure the highest levels of data accuracy and quality, when AI accesses core systems it will be subject to the old IT adage of “garbage in, garbage out.”
Connectivity and access to the right data at the right time go hand-in-hand if insurers want to make decisions that deliver real value to their customers.
Fourth, where is your human-in-the-loop?
Although AI is often depicted as a Nemesis for our species, the truth is that insurers will always need people to make smart decisions, ask the right questions, weigh query results, identify anomalies, understand fellow humans' fears, and decide how to enact strategy.
An essential shortlist continuum requires that there is non-negotiable human input in:
1. the appropriate training of the AI
2. the initial ideation process
3. checkpoints before execution
4. surveying the results to make sure the system is operating correctly
IT history should remind us that the advent of the web and e-commerce disrupted some giants and did great damage to new and old organisations that were early adopters of technology but didn’t think carefully about strategy and change.
Insurers should not forget that fact, but will need to move quickly anyhow.
We feature the best AI website builders.
This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.
The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit