Formalizing a data ethics framework can help provide a model for how an organization should act when it comes to ethics. The trick to building that framework is understanding the values the organization stands by, explains Helena Schwenk, VP of Exasol’s CDAO Office.
It’s not surprising businesses are starting to sit up and take note of data ethics.
The headlines provide a stark reminder of why it’s important. Whether that’s certain customers paying much higher insurance premiums, AI hiring systems favouring white men, risk assessment tools sending the wrong people to prison or credit card companies giving lower credit ratings to women.
Any organization using data and automated decision-making needs to think seriously about the potential for unintended consequences and disadvantaging, discriminating or harming individuals or certain groups.
When having those all-important conversations around data ethics, a fresh approach is needed. One where the discussion moves from the philosophical to the actionable, where the starting point is values-based and answers the question ‘what are our core values as an organization?’. This provides the north star to guide your organization.
Value driven data ethics places an emphasis on having a common understanding about the values the organization stands by when using data. It encapsulates its way of working, its governance frameworks and the behaviours that guide how the organization can be a data centric and ethical business. It’s this clarity between values and data centric behaviors that brings data ethics to life, makes it actionable, and far less abstract.
For example, when thinking about providing data driven recommendations for products and services to customers, are you a customer centric, high touch organization or are you providing a cost-efficient, value for money service? Aligning a data use case with values is critical to help guide decisions around data usage.
Making the explicit link between organizational values and data ethics also helps prevent another common misconception. The erroneous assumption that data ethics is primarily a privacy and security compliance exercise. Too often we see businesses viewing data ethics as ‘complying with rules’. It needs to be much more than that. It should balance the use of data – looking at the benefits, opportunities and risks – with ethical use. In short, being GDPR compliant is not a proxy for data ethics. And put another way, tor data ethics to be practical, it needs to move from ‘what not to do’ towards ‘what is the right thing to do?’.
Formalizing an ethics framework can be a helpful way of providing a model for how to behave ethically, offering a principles-based approach to doing business. It can lay the groundwork for underlining how data is used to inform decisions and ensure that data ethics practices are present in data touch points across the business. More importantly, it can become a guide to handling those difficult grey-area situations where ‘doing the right thing’ is difficult to assess or interpret.
Having a value-based approach will set you up for future data success. But it is only a starting point. A data ethics practice ultimately needs to be a multi-faceted endeavor that blends risk assessments, value creation, strong leadership and oversight, trust and transparency alongside robust data management processes. Ultimately it should imbue a solid understanding of data ethics, and ensuring that runs through the data life blood of the business.
Helena Schwenk is VP of Exasol’s CDAO Office. Helena has more than 26 years’ of experience working in the data analytics field, having spent 18 years as an industry analyst specialising in Loading...big data, Loading...advanced analytics and AI.