Panellists at an Insurance Times webinar highlighted that although artificial intelligence and machine learning will increase within the insurance sector, there will need to be governance introduced around it to ensure ‘the machines know what they’re doing’
In order for artificial intelligence (AI) and machine learning (ML) to fully flourish within the highly regulated insurance environment, governance will need to be introduced “to make sure that the machines know what they’re doing” said Greg Brown, partner at Oxbow Partners.
Brown was speaking at an Insurance Times webinar, sponsored by SSP, on 30 July 2020; the virtual session was titled ‘The legacy of Covid-19 will be the end of legacy – how a global pandemic is accelerating the adoption of cloud platforms’.
Discussing technology and AI trends that are driving efficiencies in insurance processes, Brown told webinar attendees about technology organisation Tractable, which uses AI “to analyse photographs of motor damage and then assessing the repair costs of that”. He added that AI is also being used to combat insurance fraud.
However, “quite a lot at the moment tends to be on niche examples rather than across the value chain” he said. “I certainly think that it will be the niches to start with, but it’s going to be a long journey. It’s still very early days.”
As the use of AI and ML increases, Brown identified another point for consideration – how this technology will fit in and adhere to the insurance sector’s heavily regulated environment.
He continued: “You quite quickly get to the point where the regulation requires that these things are auditable and therefore if I’m using AI and machine learning, which in its purest sense is you just let it go and it runs and it learns.
”The challenge with that in insurance and a highly regulated environment is the governance needs to be around it to make sure that the machines know what they’re doing, otherwise you end up with a situation with Microsoft chatbot and your machine becomes racist in 24 hours.”
Eastern advancements
Jasvinder Gakhal, managing director at Direct Line for Business, added that locations such as Singapore, Hong Kong and China are hotbeds of technological innovation when it comes to insurance.
“They had video tech so you could send a video or a photograph of vehicle damage and they were able to assess from the photograph and video how deep it had gone or whether it was just surface damage or whether it had gone to the engine,” she explained.
“They were getting to the point where they could then assign that to your local garage, so they knew who had parts in stock, who didn’t have parts in stock.”
Following on from Brown’s point on fraud, Gakhal highlighted technology such as facial recognition being used to determine application fraud.
She said: “You record a short video of yourself giving a description of what you need the loan for, a bit about yourself, and they were able to use facial recognition technology to decide whether it was a fraudulent application or not a fraudulent application with an 80% plus success rate.
“They were able to use facial recognition instead of password entry because they found the number of unique facial points.”
She added that technology in insurance processes across Eastern countries was “really, really progressive”.
One platform
As the insurance sector becomes more accepting of technologies, partly in response to the Covid-19 crisis, Kevin Gaut, chief technology officer at SSP, told webinar attendees about the “one platform approach” that SSP is currently using within its own cloud-based work.
He explained: “We see that the traditional roles of what an insurer does and what a broker does are merging.
”We’ve seen and had conversations with insurers who start to want to run a panel and we’ve got brokers who have started their own MGAs and so we see that these lines are blurring.
“Any platform needs to be extensible and needs to not just think it’s the thing. It’s not a new mega policy admin system. What it needs to do is it needs to sit at the heart of a wider ecosystem.
“If people want to build apps on the front end, people want to build their own interactions, people want to integrate their own systems or create systems, then absolutely the platform should be able to do that seamlessly with minimal effort.
”But equally, it needs to do the things that it needs to do in [the] best possible way. So, that’s for us what we see as a platform. It’s the ability to take a single set of components, configure those to meet the needs of our customers, be completely open and transparent as to how our customers want to use that, whether they want to use the capabilities or use something else, and that’s how we see one platform.”
The webinar also featured insights from Jim Sadler, chief operating officer at SSP, and the event was hosted by Insurance Times’s content director Saxon East.
The legacy of Covid-19 will be the end of legacy – how a global pandemic is accelerating the adoption of cloud platforms
- 1
- 2Currently reading
Artificial intelligence will need regulatory governance for use in insurance processes
- 3
- 4
No comments yet