Page 35 - BIPAR Annual Report 2020_EN

Digitalisation

The consultation paper proposed changes relating to the following topic areas: Big Data and automation in financial advice EIOPA has also identified some risks that can arise
if firms do not have in place adequate governance
- Illustrations of what the retail investor might receive in return from their investment (performance scenarios: Background arrangements. Such risks, which are not new, but
future performance scenarios, the inclusion of past performance, …); their significance is amplified when using BDA, are:
- Information on what the costs of the investment are; Big Data is defined by the Joint Committee of the European
- Specific issues for different types of investment funds (in particular the case of UCITs and relevant non-UCIT Supervisory Authorities (ESAs: EIOPA, ESMA and EBA) as - Ethical issues linked with fairness of BDA
funds); “situations where high volumes of different types of data use and the principles for responsible behaviour in
- Specific issues for PRIIPs offering a range of options for investment (so-called “Multi-Option Products” or produced with high velocity from a high number of various a digital age.
“MOPs”). types of sources are processed, often in real time, by IT tools - Other issues linked to accuracy,
(powerful processors, software and algorithms)”. EIOPA transparency, auditability, and explainability of
BIPAR responded to the consultation, stating that the legislators and supervisors should take the time to carry out a launched in June 2018 an EU-wide thematic review on the use certain tools such as artificial intelligence and
proper review of the elements which have shown to be incorrect. In the meantime, the KIDs (or elements of the KIDs) of Big Data by insurance undertakings and intermediaries. machine learning. For example, certain outputs
that are considered to be misleading should be removed by the supervisory authorities. The sector needs regulatory This review came as a follow up to the ESAs cross-sectorial reached by “black box” algorithms cannot be
stability. Frequent changes or “clarifications” to legal texts do not help creating trust and confidence by consumers and review of the use of Big Data by financial institutions published adequately explained raising questions about the
are very costly to the entire economy. in March 2018. accountability of those firms using them.
- GDPR-related issues; it is questionable
As part of the review, the European Commission, in cooperation with the ESAs, also undertook a consumer testing In May 2019, EIOPA published its Thematic Review on the use whether consumers are fully aware of how their
exercise to assess the effectiveness / retail investors’ preferred option regarding performance scenarios and past of Big Data Analytics (BDA) by insurance firms and insurance personal data is being used when they accept the
performance information within the PRIIPs KID. The results were published on 27 February 2020. intermediaries focused on motor and health insurance. The terms and conditions of their insurance policy
review revealed a strong trend towards increasingly data- and also how firms can explain to consumers in a
driven business models throughout the insurance value meaningful way the functioning of BDA tools in the
Next steps chain context of “black box” algorithms, when complying
with GDPR requirements.
The ESAs intended to conclude their review around the end of the first quarter of 2020 and submit their final proposals According to EIOPA, there are many opportunities arising - Issues linked to personal data and
to the European Commission shortly afterwards. This has not yet been done. from BDA, both for the insurance industry and consumers: consumer protection regulations when the
information is stored on the cloud. Also, an excessive
The European Commission’s wider review of the PRIIPs Regulation (level 1) will also look at the interaction with other - The combination of traditional and new sources concentration in the number of providers in certain
recent new information requirements such as in IDD and MiFID II, how they work together and if the interplay can be provides greater granularity and frequency of information strategic services/technologies can potentially
improved. about consumer’s characteristics, behaviour and lifestyles. disrupt the efficient functioning of value chains,
This enables the development of increasingly tailored leading to situations of “reverse outsourcing”.
The Commission is expected to soon launch a study to bring all of this together. products and services, more accurate risk assessments,
speed in decision making and subsequently less operations EIOPA’s work on digital ethics in insurance
costs. In 2019-2020 EIOPA conducted further work on
- BDA tools such as artificial intelligence and machine ethical issues with the fairness of the use of BDA,
learning are commonly used by insurance undertakings as well as regarding the accuracy, transparency,
rather than by insurance intermediaries, primarily on auditability, and explainability of certain BDA tools,
pricing, underwriting and claims management. such as artificial intelligence and machine learning in
- The use of cloud computing services is increasingly collaboration with the industry, academia, consumer
widespread, enabling implementation of BDA solutions. associations and other relevant stakeholders.
Uptake of usage-based insurance products will gradually EIOPA also continues its ongoing work in the area of
continue in the following years, influenced by developments cyber insurance and cyber security risks (see under
such as connected cars, health wearable devices (Internet of Cybersecurity for more information).
Things -IoT) or the introduction of 5G mobile technology. This
can help consumers to obtain a more accurate calculation As a follow-up of this thematic review, EIOPA
of their insurance premium. Robo advisors and specially established in September 2019 its Consultative
chatbots are also gaining momentum. Expert Group on Digital Ethics in Insurance. The
objective of this Group of experts is to assist EIOPA

34 35