There are a myriad of ethical issues circulating around the insurance sector in relation to big data. So how can you organise those ethical issues in a way that allows for a more strategic response. This framework will get you started.
First off though, you need to address why it’s worthwhile organising your response to these ethical issues into some form of strategic framework. The answer lies in these two statistics, from a survey published by Price Waterhouse Coopers in February 2017:
- 72% of insurance CEOs think it is harder to sustain trust in a digitised market
- 28% of insurance CEOs are “extremely concerned” that trust will affect their firm’s growth
This worrying picture gives you the title for the framework for your strategic response to data ethics. It should be called ‘ Trust and Growth in a Digital Market ‘.
The framework I’m setting out here has three component to it: data, algorithms and practices.
The Three Dimensions of Data Ethics
The ‘data’ dimension of data ethics focuses on ethical problems posed by the collection and analysis of large datasets. So this would cover how the data held by an insurance firm is generated and how it is recorded. And it would cover how it is curated, how it is processed, how it is disseminated and how it is used.
The ‘algorithm’ dimension of data ethics addresses problems posed by the increasing complexity and autonomy of algorithms. So this would include initiatives such as social network analysis, machine learning, internet bots and natural language processing. At the same time, it will cover themes such as design, training, testing and auditing.
The ‘practice’ dimension of data ethics looks at the responsibilities of people and organisations in charge of data policies, strategies and processes. So this would include how executives understand their firm’s responsibilities and how they are shaping their leadership in fulfilling them. And it would also include how it is setting and communicating the principles around which it wants data innovations to be taken forward and how it is driving those responsibilities into everyday business decisions.
These three ethical dimensions of data, algorithms and practices provides an inclusive framework within which this diversity of ethical implications can addressed. Flip it over and it also provides a framework around which the ethical oversight of a firm’s digital strategy can be organised (more on this here).
Algorithm Regulation
Does your firm have time to organise and introduce such a framework? That’s the wrong question, as least in the UK. It should instead be about how the firm makes time, given the extension in 2018 of the Senior Managers and Certificate Regime (SMCR) to the insurance sector. Remember that ‘algorithmic trading’ has already been associated with a ‘significant harm function’ in the SMCR regime under which banks already operate. So can insurers assume that their use of algorithms in core operating systems will be treated differently? I think not.
Insurers need to grasp the ethical issues underlying ‘ trust and growth in a digital market ’ and organise their response. This framework provides them with a template for doing so.
This post owes much to the paper ‘What is data ethics‘ by Luciano Floridi and Mariarosaria Taddeo. Published in Philosophical Transactions of the Royal Society, Dec 2016.
