How will customers interact with insurers in the future? Is insurtech hype? What technology is already being used? What are the legal implications of these new technologies?
These questions were considered by the industry at the 2019 ANZIIF Insurtech Conference held last week in Sydney. Held in conjunction with Insurtech Australia and Insurtech New Zealand, it was a great opportunity for people in different parts of the value chain to meet and collaborate, with delegates coming from established insurers, Lloyd’s coverholders, brokers, insurance startups and other insurance service providers.
An ‘uber’ moment for insurance
Andrew Rear, CEO of Munich Re Digital Partners (MRDP) gave a keynote presentation at the conference. MRDP works with new and emerging companies to bring their product to life. It is a global venture, and generated A$150m in Gross Written Premium (GWP) through its partnership with new and emerging companies in 2018, and is set to make an estimated A$300m of GWP in 2019. MRDP can help new and emerging companies access capacity.
Andrew predicts that there will be an ‘uber’ moment in insurance. It will cause disruption but will not subvert the industry as we know it. Today’s insurance brands are not going to disappear overnight. However, they might need to spend quite a bit of money to buy out new competitors and, in doing so, will integrate new technology into their operations. He also believes that the retail consumer market is likely to change more quickly than the corporate insurance market.
What is already here?
Digital distribution with automated underwriting for personal and small business lines is already here. These distribution platforms may use machine learning to deliver an improved customer experience. Hippo Insurance is one such venture backed by MRDP. Hippo provides home insurance in the USA and can provide a quote within 60 seconds. It does this by using big data and artificial intelligence (AI) to analyse and integrate publically available property information. For example, AI can review aerial imagery to determine the condition of the property and proximity to bushland.
AI can also be used to enhance the customer experience. Dr Catriona Wallace, founder of Flamingo AI gave a keynote presentation exploring the ways AI is already changing the sales process. Instead of taking customers through pages of forms, virtual assistants powered by AI can guide customers through the process from quote to binding of a consumer insurance policy. This may be fully automated or human assisted. For example, a staff member may join the conversation if the virtual assistant encounters an unfamiliar question. The AI is able to learn from this interaction so that it is able to handle similar situations in future.
What is coming?
Blockchain based policy administration is a potential new technology. However, there are differing views as to whether blockchain based policy administration is necessary in the insurance industry. On one view, blockchain acts as an automated single source of truth, which can be useful when dealing with members of the public such as on a cryptocurrency exchange. However, insurers are generally well-established entities and are unlikely to disappear overnight. The benefit of having a publically available single source of truth may not be of substantial benefit.
Another new technology on the horizon which will impact claims departments is automated loss adjusting. Although insurers are increasingly using drones and mobile technologies to undertake virtual loss adjusting, the damage still needs to be assessed by human claims officers. A possibility in the future is for computers to be able to assess the damage automatically with little or no human interaction.
Also on the automation front, autonomous vehicles will become ubiquitous in the medium term. Alex Taylor, IAG’s entrepreneur in residence, considers that self-driving cars will change ‘societal fundamentals’. One example that he gave, quoting Goldman Sachs, was that self-driving cars could save the freight transportation industry in the USA $168bn annually but lead to a rise in unemployment by 7% due to the ‘cascade-effect’ loss of jobs in not only the driving industry, but also service industries that support trucks and truck drivers. Alex considers that self-driving cars have the potential to dramatically change insurers’ businesses with a reduced accident rate of up to 90% putting downward pressure on motor premiums. Motor insurance is a significant component of local insurers’ underwriting books.
What are the legal implications?
Here on the Insurance Law Tomorrow blog, we like to consider the legal ramifications of the changes to the industry. As discussed during the panel discussion at the conference, establishing a partnership between new and emerging businesses and established insurers can be challenging and subject to power imbalances. With many insurance startups needing access to data, the sharing of data and responsibility for compliance with privacy laws must be addressed. This is particularly important given Australia’s implementation of a Notifiable Data Breach scheme, which has now been operating for just over a year.
The use of AI in the insurance sales channel, for example through a virtual assistant, is also of interest. Insurance is a highly regulated industry. An Australian Financial Services Licence (AFSL) is generally required to sell insurance to retail customers. There are also restrictions on providing any sort of financial advice relating to general insurance products. These measures, in place to protect consumers, are only set to become more stringent following the recent Royal Commission into Misconduct in the Banking, Superannuation and Financial Services Industry.
The insurance sales channel for consumer insurance often involves the provision of general advice. General advice is a recommendation or opinion that does not consider the client’s objectives, financial situation and needs. An AFSL is required to provide general advice, unless an exemption applies.
ASIC’s Regulatory Guide 255 provides some guidance on provision of digital advice to retail clients. Although it does not specifically relate to AI and virtual assistants, it does address advice which is given using algorithms and technology without the direct involvement of a human adviser. An AFSL holder must have adequate resources to comply with their obligations. In relation to digital advice platforms, ASIC expects there to be people within the business who understand the rationale, risks and rules behind the algorithms used to generate the digital advice, and who can review the digital advice.
ASIC also expects AFSL holders to monitor and test the algorithms used to generate the advice, including having appropriate controls and procedures to monitor and supervise these algorithms. If the digital advice platform is outsourced, the AFSL holder remains responsible for the platform and ASIC expects the AFSL holder to still have knowledge of the rationale, risks and rules behind the algorithm, even if they may not understand the specific computer coding.
The insurance industry is always adapting to new technology. However, as technologies become more complex, companies should make sure that they understand how they work and comply with regulation in this area. Ultimately, the aim of regulation is to ensure consumers are able to make informed decisions and generate positive consumer outcomes, which should be part of the value proposition of any insurance product.