Does the data industry need a code of ethics?
When it comes to Doing Data Right, Gillian Docherty has a simple mantra: “Just because you can doesn’t mean you should.”
The chief executive of The Data Lab is well aware of the enormous potential in the field. As its website says: “We believe Scotland can lead the world to a future where data powers scientific progress, economic prosperity and social good.”
Docherty believes all three objectives can work together in a positive way, but must be accompanied by a constantly questioning attitude to how and why data is being deployed.
“As capability in this area increases and organisations across the public and private sector do more with data, you have to bake into the heart of that growth the question, ‘Should we do it?’ – whatever your business priorities and imperatives.
“If the Cambridge Analytica and Facebook data scandal taught us anything, it was that they did not ask that question, ever.”
Docherty says the whole area of data ethics is still emerging: “It’s where all the hype is now – it used to be big data that everyone talked about, now it’s data ethics. It’s fundamental, and embedding it across an organisation will give competitive advantage.”
So what is The Data Lab, set up in 2015, doing itself in this ethical space? “We’re ensuring data ethics training is baked in to the core technology training of all Masters students, so they are asking all the right questions,” says Docherty. “The number of Masters studentships has grown from 40 in our first year to 155, across 11 universities – that will have a big impact over time.
“In many ways, it already is having an impact because candidates ask ethical questions about businesses they might be working for. We have seen pushback from Google employees about the artificial intelligence used in certain projects, and we are hearing from many corporate recruiters that candidates look at the company’s whole approach to ethics and sustainability. They want a value system and data ethics is part of that.”
Docherty recognises the pace of technological change is a huge challenge in data ethics: “There are times when the business imperative?pushes you to do something that in hindsight you would perhaps not have done. You need to align values with ethics to give you capability. If you align everything, there might be projects you decide not to take?forward because they aren’t right.”
Alongside an ethical framework, do we need stronger regulation too – and can that ever keep up with the pace of change? “It’s clearly a challenge,” says Docherty. “One of the conversations is whether data scientists should take a kind of Hippocratic Oath: ‘At first, do no harm.’ We might well see frameworks like that ?emerging as a form of self-policing and self-regulation.”
What about state-led legislation and regulation? While the European Union created the General Data Protection Regulation (GDPR) to provide a framework for dealing with data, some describe attitudes in other parts of the world as ‘The Wild West’. As Jarmo Eskelinen points out in his introduction, the ‘deep state’ approach in China and free market attitude in the United States are hugely different to what Europe has tried to achieve in the data space.
“When we are working in a global market-place, with no barriers or boundaries, legislating and regulating is very difficult – because ?legislation and regulation inevitably focus on a specific jurisdiction or ?industry,” says Docherty.
“The lack of common regulatory and ethical frameworks is a huge problem in a global industry without borders. Those frameworks are emerging and will be increasingly important, but there are hundreds globally and it’s very difficult to align them with our values and organisations. They need to be built in to what organisations are doing with data, not added as an afterthought.”
Docherty sees huge opportunities for Scotland from this fragmented regulatory and legislative landscape: “I think there is a tremendous opportunity because of Scotland’s reputation for its culture, governance and progressive approach; this could help deliver a genuine global leadership role in doing data right.”
Linked to the ethical question is whether advances in data science should be more focused on doing public good. When it comes to The Data Lab’s three key areas – that data power should drive scientific progress, economic prosperity and social good – some critics argue the social good element can be downplayed. Does Docherty agree we should harness data’s power more effectively for social benefit?
“I think there is so much more we can do with data across the board in all three areas of focus, to drive scientific, economic and social benefits. There are so many opportunities to use data in new ways and there is a lot of work going on with senior leaders across different sectors as no area has it totally cracked. Do we need to do more around social good? Absolutely. Do we need to do more on ?economic benefit? Yes.
“I think we are only scratching the surface of the opportunity. Some organisations are more advanced than others but there is no organisation, public or private, that could not do more with data.”
Fit for purpose
One project that cuts to the heart of Doing Data Right is the FIT Home initiative in the Highlands. The project also illustrates the use of data for social good and the ability of the public, private and academic sector to collaborate effectively.
The FIT Home initiative involves Albyn Housing Society, in partnership with Carbon Dynamic (a manufacturer of sustainable, modular homes) and NHS Highland. It aims to design homes to enable people to live independently for longer – providing a secure environment for them to live and focusing on keeping them well and at home for as long as possible.
The Data Lab provided funding to involve Robert Gordon University in the project. Professor Susan Craw and her team used their expertise in artificial intelligence to use specialist sensors (FITsense) and recognise residents’ activity from the data collected. From this, they determined patterns of activity (including movement and everyday living) and, specifically, any changes in patterns likely to lead to dangerous falls.
Examples of FIT Home features include flexible walls and spaces for storing additional medical equipment, to allow residents to be treated at home instead of in hospital. The project aims to reduce pressure on public services and resources, by predicting and preventing injury. With injuries from falls costing NHS Scotland £471 million a year, the wider economic benefits of the scheme could alleviate pressure on public services – and empower residents, care providers, social landlords and health professionals.
The FIT Home technology has the scope to be retrofitted into any home. The long-term plan is to open the technology to commercial availability within a social business to support an increasingly ageing population.
Data Lab chief Gillian Docherty says: “This project involves a really interesting group of collaborators and has allowed people to live independently – and to approve the sharing of their data and to view it on TV screens in their home. A huge amount of work went into the ethics and data space in that project – and a lot of time spent ensuring residents were comfortable with everything.
“This project is about collecting data for a very specific, positive purpose. When you do things right, with sensitivity, and when you embrace privacy issues, everybody gets value.”