SIRC: Harnessing machine learning, AI and data governance
November 3 2020 by Andrew Tjaardstra
Russell Higginbotham, chief executive of Reinsurance Asia at Swiss Re, introduced a high level digital panel discussion on November 3 at the SIRC 2020 virtual event.
He said that the pandemic has helped transform digitalisation, but that it’s very important for the market to adopt technology sensibly and appropriately.
Mega trends
Moderator Pranav Pasricha, global head of P&C solutions at Swiss Re, noted the market is seeing several big digital mega trends.
“We look a new business models: intelligent insurance – more active products – usage-based insurance – products that put consumers at the heart. It has become more about claims, security and peace of mind, and more dynamic business models,” he said.
“Who is liable for artificial intelligence when AI makes a mistake?” Pranav Pasricha, Swiss Re
Pasricha added: “Another trend is autonomous processes – applying data sources to things that will operate better than humans – such as automated underwriting, no touch claims and parametric coverage.”
Meanwhile, the value chain is evolving. He noted that Tesla is becoming a carrier for their clients, which presents a market opportunity.
There are also new risk pools, such as non-damage business interruption, and Pasricha posed this important question: “Who is liable for artificial intelligence when AI makes a mistake?”
It’s harder for smaller firms to innovate, added Pasricha, highlighting Swiss Re’s time spent on risk consulting and ground-up product development.
Shift in value
Digitalisation has also allowed insurance to be increasingly unbundled.
Arijit Chakraborty, Asia managing director at insurtech firm Cover Genius, commented: “We are shifting the value in distribution. And we have unbundled insurance — you can use AI, machine learning, to help facilitate this. For instance, bite-sized life and health plans when opening a bank account and (we have also created) logistics insurance for exports.”
Chakraborty explained: “Insurance is becoming more important and it’s easier to give customers a journey. We have redrafted all the policy documents to make it easier for customers. We look at the seasonality and when people buy, as their risk profiles are different.”
“There has been a shift in claims management. You can go into far more detail – going into losses in particular cities and personal profiles,” he added.
“There is no silver bullet for digital.” George Kesselman, ZA Tech
Another panel member, George Kesselman, head of commercial for ZA Tech and president of the InsurTech Asia Association, commented on the future of digital: “There will likely be a hybrid model going forward. Meta-data around the risk is also becoming extremely powerful.”
But Kesselman added a caveat: “There is no silver bullet for digital.”
He explained that the digitalisation of insurance is largely following the same trends in other markets and still in the ‘finding scalability stage.’ Once insurance has passed this stage, AI can become more powerful — in applications for behaviour modelling and meta-data, which will bolster the wider market.
Data
As the amount of data grows exponentially, data governance is set to become far more important.
Ronald Chung, managing director of data trust services at PwC, commented: “Data governance needs to change because of digital and AI. There’s more data being captured — but is the data being governed at all?”
“There is also the problem of data leakage and we have seen some high profiles recently in in Singapore. You need to maintain old databases across multiple storage locations.” Ronald Chung, PwC
Chung noted insurers aren’t as heavily regulated as banks, but digital and AI use a lot of data. The question of where the data comes from and if the market is using the right quality of data, also comes into play.
He also warned: “There is also the problem of data leakage and we have seen some high profiles recently in in Singapore. You need to maintain old databases across multiple storage locations.”
Working with an agency workforce that created a digital app to capture more data, Chung realised you need to incentivise agents and give them reasons to collect more data — and for the customers to provide it.
It will also be important for firms to upskill and for employees to understand how to use automated tools. While there is often push back from staff because people feel they may lose their jobs (due to incumbent tech), he stressed the importance of automation to the upskilling process to carry out the jobs for the future.
Luckily the Singapore community has understood this and is ahead of the curve in comparison to many other markets.
InsuranceAsia News is a media partner to the SIRC 2020, which is taking place digitally from November 2 to November 5.
-
BHSI | Managing non-Asian exposure in long-tail lines
While US-exposed business can look attractive to Asian carriers, managing the volatility around the long-term results and the ability to model those losses are crucial, say BHSI’s Marc Breuil and Marcus Portbury.
-
Sedgwick | To Handle CAT Claims Well, Multi-Step Preparation is Key
When it comes to risk, it’s not a matter of “if” it’s a matter of “when” an event will occur.
-
HSBC Asset Management | Is it time to relook at Asian currency bonds?
With diversification and performance high on investors’ agendas, it seems a good time for global portfolios to revive allocations in Asian local currency bonds – including Hong Kong dollar (HKD) bonds.
-
PineBridge Investments | Why Asian insurers are exploring private credit and CLOs
The recent rollout of risk-based capital regimes across Asia calls for a closer alignment between insurers’ assets and liabilities. We explore potential ways to maintain a healthy investment yield and robust returns on regulatory capital.
