Translating data ethics
We’re sharing our work on permissions and transparency with other product teams whose first language isn’t English, so we can improve the principles, test them and make sure they’re valid. We’ve translated our data ethics principles into Arabic, French, Spanish, Mandarin and Portuguese. German and other languages are planned soon, and we hope to receive feedback from people and teams that use them. We've created posters with a white background, so people can experiment with printing them out on paper in different colours. If you spot something that’s not quite right with these translations, write to us at email@example.com.
IF’s data ethics principles came from three years’ work on knotty data challenges for organisations and digital services. We wanted to codify some of the commonalities we’ve seen in product teams that are designing, making, testing, shipping and maintaining digital services that can be trusted. The principles are a useful way of thinking through data practice: they can help organisations assess their own position, and can point at and explain a service’s ethical promise.
Ethics don’t stand still
The term 'principle' might sound like a reliably unchanging law, but we don’t think of the data ethics principles like that. Ethics are a hard problem. Ethics don’t stand still. Some will need to be adapted, added to, to make sure they stay relevant as technology changes. Our principles won’t be right for every service or every data ethics problem that a product team might encounter.
Data practices and the real world
We want the principles to represent what ethics mean to product teams. They should operate at a high enough level that they can change, while also being specific enough that they can connect to real products and real practices. We expect organisations to mould the principles to fit the context - the organisation, people using the service, and data about those people. And we hope that these principles will be informed and improved by the product teams that use them.
All of this means that we expect these principles will change in the same way that software gets updated and changes.
We also expect new data use cases and challenges to emerge. For instance, how can someone see that a service’s ethical promise has changed over time, or how it has changed?
Translation isn’t just about words
We do most of our work in English, but we’re very aware that language and culture also affect what ethical practice means. Consent, understanding, transparency and accountability aren’t necessarily identical concepts applied identically everywhere.
We want to have a much broader conversation about what ethics means in countries we’re less familiar with. That’s important because we think we’ll find better examples, and we’ll also find more robust answers to what consent means in different cultural contexts.
We’re also questioning the idea that service design and data are solely about an individual and their interactions with the service. What happens if you think about data use, permission and agency in terms of a community? We’ll explore these issues and we may see entirely different models of understanding emerge.
Creating trusted services
So data ethics principles are not a set of commandments, nor a box-ticking exercise. Instead, the aim is being better guardians of data and creating services that can be trusted. We think of the principles as tools for helping teams look at, and think through, their data practices. A part of working out how to increase trust in what they build and maintain.
We’ve started offering Data Ethics Toolkits for organisations that want to use data in a ways that are ethical, practical and creative. Write to us at firstname.lastname@example.org if you’d like to hear more.