26 May, 2023
Your business has (we hope you have!) a Data Protection Officer, and data privacy policies that comply with GDPR and that ensure that you are looking after personal data responsibly and effectively. But what about the way in which you – legally – use that data? How do you use data from other sources, including AI technology, to drive decision making? Are you using biased data sets that may create biased decisions? Are you thinking about privacy, transparency and accountability when you use data? Are you telling people how you are using data to make decisions about them? Are you inadvertently sharing confidential IP when you upload data to ChatGPT?
If you aren’t thinking about data ethics, as well as data responsibility, you really should be. Given the impact your use of data has on your stakeholders and the consequences to the business of doing it poorly, data ethics is as much a material sustainability issue as D,E&I or carbon management.
While few of us are data ethics experts and even fewer of us understand much if anything about how large language models such as ChatGPT work, we need to face up to the changes that are taking place, because they have profound consequences we don’t yet fully understand, but that we know will affect us all, as consumers and as business leaders. As the very good, and very useful, new Good Data Guide from data experts Profusion and law firm Pinsent Masons points out, technology moves faster than the law, so the Guide – which sets out to provide a practical toolkit for all aspects of data ethics – will be updated regularly.
The guide emphasises the importance of culture and of making this everyone’s responsibility. Education is key, as is recognising that while technology can do things we are simply unable to do without it, in the end the moral responsibility for making good choices rests with us. If we want the trust of the people we interact with, and we want to use data in ways that improves rather than holds us back, these good choices are essential. The guide emphasises principles and simplicity, and the importance of building on existing policies, rather than building new ones. As such, any organisation that collects and uses data – which is essentially every organisation – will find something of interest in the Guide.
By Claire Jost