“Ethics and innovation are not mutually exclusive. Thinking carefully about how we use our data can help us be better at innovating when we use it,” writes the then Secretary of State for Digital, Culture, Media, and Sport on the government’s official ethics framework webpage.

This government-level focus on the ethics of data isn’t just relevant to those who work in the civil service. Instead, any organisation that handles data or conducts data projects can look at what is outlined in this new framework and learn from it.

What is the story behind the new framework?

This new ethics framework aims to lay out best practices for the use of data in the public sector and is a timely update to an existing policy (from 2016). Outdated ideas of data is a pertinent problem for the government to tackle since stories about data breaches continue rolling in from across the public and private sectors. For instance, the misuse of personal data collected from Facebook for political and advertising purposes, a Dixons Carphone data breach that compromised 105,000 payment cards, and the recent NHS data breach that affected 150,000 patients across England are just three examples of recent data issues.

The old policy is no longer functional as advancements in data acquisition and usage are changing. It’s also interesting to note that the announcement was made during London Tech Week, which implies that the UK government believes that the framework can benefit more than just individuals; instead, the framework will become a crucial part of UK’s future data landscape in both public and private sectors.

What is included in the new Data Ethics Framework?

The framework consists of three parts: the data ethics principles, additional guidance for each principle, and a workbook for teams to apply the framework.

The seven principles are as follows:

  1. Start with a clear user need and public benefit
  2. Be aware of relevant legislation and codes of practice
  3. Use data that is appropriate to the user need
  4. Understand the limitations of the data
  5. Ensure robust practices and work within your skillset
  6. Make your work transparent and accountable
  7. Embed data use responsibly

How other organisations can apply the framework

Businesses can apply many of the same ideas in internal data frameworks. Here’s how:

Create a data framework

In creating a framework for data science ethics, the government has demonstrated something important: make sure there is an agreed, standardised practice for everyone involved to follow. In having a framework, there is an agreed set of guidelines that gives everyone something to refer to if data issues or questions arise. It also helps in looking back if anything goes wrong with a data science project.

Evaluate existing work against the framework

One easy way to use the framework is to compare the government’s principles to the organisation’s own principles, or to the standard of work the organisation has already done. Organisations can pick one project or one area of their businesses where data features heavily and test it against the framework, starting from point one. As they move through the principles, they can make sure to give the example an honest evaluation.

In using the framework as a testing tool, businesses can learn a number of things: one, whether any existing framework truly worked; and two, how much work there is to do in improving future projects.

Understand and apply its key ideas

The framework seems simple at first glance, but there is a lot of detail that goes into each of the principles. This can be overwhelming; however, organisations can benefit from looking at its fundamental ideas and tackling those first.

If organisations want to start from somewhere, there are two areas the framework is aiming to get across amongst all others.

Firstly, transparency and openness. Organisations should think about the example they use in their deeper explanation of principle number six: ‘‘…scientists share data when publishing a paper on Figshare and Datadryad. This gives others access to the data and code so that the analysis can be reproduced.”

‘Responsibility’ is also a keyword in the framework, particularly on the appropriate use of data once it is collected. This point breaks down the responsibility in adapting a policy with data, data maintenance or performing a redesign of a data model, as well as the ongoing responsibility to keep within the framework guidelines.

This is a responsibility all organisations need to understand. Where does this responsibility lie? If it is with the individuals themselves, then skills and training need to be offered. If the responsibility falls to a certain set of people, they need to be aware of that role.

Processes and the ability to ask questions and flag issues are also key to the framework. Would staff members feel comfortable flagging data as flawed, or biased? Creating a standard process, in this case, can eliminate any caution about highlighting problems.

Adopt its practicality

The creation of a workbook is another crucial element to take from the framework. Laying out a framework and instructing staff is one thing, but giving staff a practical way to apply the framework to their projects ensures that it is being put into practice.

Skills and training are an important addition here. If members of staff at any level involved with data are unable to understand general data science ethics, and how the organisation intends to apply it to their work, then they will fall at the first hurdle.

Conclusion

The framework from the UK government is not applicable to the civil service alone. In producing the framework, the government has already done a lot of the work for organisations that work with data as well. Organisations can use this as a launchpad for developing their own data science ethics framework. They have revealed the pressure points, the key themes of data science ethics, and given everyone something to use moving forward.

Econsultancy reports:

Econsultancy training:

The Festival of Marketing 2018, 10-11 October in London, features a Data & Analytics stage (one of 10 stages).