Embedding data ethics in an organisation

Data Ethics isn’t just something that companies should be conscientious of, it should be a central to how their culture, values and how they operate. It is important, if not critical to the bottom line of modern business. This was the topic of the ODI’s most recent Canal Side chat, which saw the ODI’s Head of Consultancy for Data Programmes, Lisa Allen, in discussion with Kevin Macnish, a Data Ethics Consultant from Sopra Steria.

It was put forward that investors, customers and employees seek companies that are ethical when it comes to how they collect, store, and use data. That future generations, particularly Generation Z put more value on ethics, as well as sustainability and personal values then other generations. And how our education system and government need to do more to teach and govern ethics when it comes to data (it was optimistically estimated that only around50% of IT students study ethics during their course).  

The key takeaways were that businesses need to think hard about how to embed a culture of ethics around the full data life cycle of their business. And that ultimately, it’s up to leadership to ensure this happens.

Reflecting on this point about leadership, the real challenge is addressing the gap between the idealistic state a data ethics culture and the operational reality that companies face in creating and instilling one. There are great frameworks available to help guide discussion around data and ethics(such as the ODI’s Data Ethics Canvas), but there is still a lot of ambiguity around what is meant by ethics, how to align employees around ethics, how to have an open, progressive dialogue about ethics, and how to have real buy in across the organisation towards a cultural of data ethics.

The real challenge is that data ethics can’t just be a top-down directive. By being too dogmatic, leadership can risk putting people off having open and honest conversations and can create a culture that discourages exploring innovative ways to use data to drive an organisation forward. As with any cultural change the real challenge is making data ethics a part of the fabric of everyday operations in an authentic, collaborative and meaningful way(which the ODI’s Data Ethics Canvas does help to do) rather than through additional policies, bureaucracy and policing.

A few activities which can help to do embed data ethics as part of company culture are:

1.    Create your own framework that works for your organisation

There are a lot of great data ethics frameworks out there which can help you assess and explore data ethics throughout your organisation. It’s a good idea to explore different frameworks to see what works for you, but then consider developing your own approach to assessing, monitoring, and embedding data culture within the context of your industry and your organisation. By creating a tailored approach can provide some contextual focus on the specific topics, data and processes that are more relevant to your company as well as alignment with your companies’ core values and objectives.

2.    Apply the framework to your existing data lifecycle and ecosystem

Map out your data lifecycle and use the framework to critique each process, activity, and technology involved in data collection, processing, storage, and usage. Invite relevant stakeholders to participate the parts of the data ecosystem that are relevant to them. Look a tall areas where that could be impacted – from internal operations to external audience engagement and any  parties you work with. This is a great opportunity to engage with teams across the organisation with interactive workshops where data ethics can be explored within the context of their day-to-day work.

3.    Set up a data ethics community

Consider creating a data ethics community bringing together people across the organisation who have an interest in data ethics. This can be a great way for different people to openly discuss and debate data ethics as a broader concept, to voice opinions about how your organisation’s stance, and to surface new ideas about how to become more ethical. Task the community with devising ways to promote data ethics throughout the organisation, to organise events around data ethics, and explore new tools, frameworks and partnerships that may help the organisation to adopt a culture of data ethics.

4.    Run regular workshops and refresher events

Encourage workshops where any new project, process, activity, or technology where data is being generated or used can be critiqued against your own or a third parties’ framework. As organisations are constantly evolving, run quarterly or bi-annually events to discuss, debate, review, and update your framework and assess how data ethics has evolved within your company. Invite all new joiners to participate in data ethics training and hold regular refresher sessions to ensure all employees have the knowledge and tools necessary to incorporate data ethics practices into their daily work.

In addition to these actions the following provide a few practical tips for exploring data ethics and what it means to your organisation:  

Just because you can do it doesn’t mean you should

There is a gap between regulation and ethics. GDPR is primarily concerned with data privacy and security, but doesn’t provide any guidance when it comes to data ethics. It’s important to think beyond existing regulations and laws, and think about what is right. For example, GDPR states that persistent cookie data can be maintained for 12 months, but from an ethical view point it may be seen as an unreasonable duration.

Think about data ethics by design

How can you embed data ethic principals when designing, building and delivering products and services? What are the tactile ways that data ethics can be operationalised beyond a framework or policy? For example, can you deploy an way to automatically check and surface any potential bias of machine learning algorithms? Or can you ensure different stakeholders have an easy way to challenge whether a new solution or product has any ethical issues?

Be empathetic towards different objectives

Although it was mentioned at the event that ethics are universal, the topic of ethics can be somewhat subjective and dependent on your role, objectives, background among other factors. For example, extending the time you keep PII data may be critical for a marketer, but seen as unethical by a developer. It’s important to appreciate the different viewpoints people will have, and encourage open conversation rather than seeing data ethics as a clear cut case.

Don’t make data ethics a mandate – invite views, debate and discussion

To get real buy in think about how you can create conversations around data ethics, encouraging debate around what does your company mean by data ethics, and how data ethics could impact different business drivers positively or negatively. By making data ethics a top-down mandate you risk stifling discussion, innovation, new idea and ultimately buy in. Inviting people to feel part of how it is defined and incorporated into their daily lives is vital to create a culture accountability and ownership around data ethics.

Don’t just think about the bad – think about the good  

When exploring data ethics, think about what good can also be achieved using data you generate or use. Are there any insights that may help drive meaningful change within your organisation, your community, or the wider world?

‍

Resources:

ODI Data Ethics Canvas and Maturity

DAMA – DMBOK

IBM data ethics

Government data ethics