Oral evidence: The big data dilemma, HC 468Tuesday 1 December 2015
Oral evidence, 1 December 2015.
Witnesses
Mr Edward Vaizey MP, Minister of State for Culture and the Digital Economy, Department for Culture, Media and Sport and the Department for Business, Innovation and Skills, Rebecca Endean OBE, Director, Research Base, Department for Business, Innovation and Skills, Baroness Shields, Parliamentary Under-Secretary of State for Internet Safety and Security, Department for Culture, Media and Sport, and Peter Knight, Deputy Director, Research and Development Directorate, Department of Health, gave evidence.
Chair
Welcome, Ministers and civil servants. Thank you very much for your time. We are very grateful to you for taking it. We are going to have a vote any minute. This is our final session on the big data dilemma. We have called it that because our evidence has been very clear that there are huge opportunities associated with big data, but also risks that are creating barriers to realising those opportunities. In the evidence from the Department for Business, Innovation and Skills, the Department was rightly proud that since 2011 the Government had invested over £500 million in developing the UK’s big data capability. While we are gone, perhaps the Ministers can think about how they are going to explain what they have achieved with that money. I am now going to adjourn the Committee while we vote. We will recommence when we are quorate. Sitting suspended for Divisions in the House.
Chair
I now resume the session. Thank you so much for waiting. You are very patient. I understand that we are also expecting a vote in the Lords. It is very unreasonable of the House to continue having votes while we are holding this very important session, but we must accept that that occasionally happens. I gave you quite a lot of time to think about my first question. What exactly have you been doing with that £500 million since 2011? Who would like to start?
Rebecca Endean
I am happy to take this one.
Mr Vaizey
I think we want to make some opening statements.
Chair
We do not have a huge amount of time, so let’s start.
Rebecca Endean
We have spent it on a range of things. I can deal with this very quickly because we are short of time. It has been invested mainly in infrastructure in the research and science base. We have spent it on high-performance computing, in particular ARCHER and the high-performance computers in Hartree; better networks for the academic network, including JANET, which connects up all the universities; better software, including the new IBM collaboration in Hartree, which brings things like Dr Watson technology into the UK; better data, storage and curation for researchers; and, finally, things like the Alan Turing Institute, which is about improving data science and making sure we get world-class scientists working on some of the key problems that face us today.
Chair
Minister Vaizey, you seem keen to contribute. The evidence from the Department for Business, Innovation and Skills states that fully exploited data could be worth about £216 billion to our economy, and other evidence says that it would create about 58,000 jobs and mean about a 3% uplift in productivity. Clearly, that is very exciting and we should be looking at how we do that, but other witnesses have told us that most companies estimate they are using only about 12% of their data at the moment and we need to look for ways to improve that. What is your strategy for doing that?
Mr Vaizey
May I start by saying that it is a great opportunity to be able to speak in front of this Committee and to give evidence with my fellow Minister and colleague Baroness Shields? You have called your inquiry the big data dilemma. I think you are quite right to balance the opportunities that big data brings with some of the risks, but what I would add is that big data is not going away. It is the Government’s role to be on the front foot. I am sure we will cover a lot of ground in questions, but, broadly speaking, it is Government’s role to be talking about big data and putting the UK front and centre of big data policy, to invest in it and to work with business. As you say, Chair, some of the calculations show that there could be a significant uplift in productivity and value. Those figures can change in real time; they may go up or down, but I think everyone would agree that if we can make progress on how companies, both large and small, use big data—not crack big data, because we will be looking at this issue constantly—and if Government can play its part in opening up data, we will see some companies benefit significantly. You said that companies use on average only 12% of their data. One of the issues we are facing is that there is, as it were, a wave of data arriving by the terabyte, and processing it is very difficult and a big challenge. People forget that there are huge datasets out there that people do not have the time, skills or opportunity to analyse. To echo what Rebecca said earlier, the Government’s role is to invest in big projects that can help companies analyse big data or invest in skills, because we need data scientists who can help companies crack those big datasets, and also work with companies to tell them about the opportunities and give them a route map to engage in big data. I am sure Joanna will want to add to some of those comments.
Chair
I will target a question to Baroness Shields in a second. To clarify one point, you said you want to make progress in the way companies use data. What would progress look like in specific terms?
Mr Vaizey
We do not have a specific figure, plucking numbers from the air, for how many companies we want to see using big data by any given date, but, for example, the digital catapult, which probably really got going last year, engages with a number of companies. We want to see that engagement increased. They have made the use of data very much front and central to their first work project in establishing themselves, but we do not have a specific figure for the number of companies we want to see engaging in this way, unless Rebecca has thoughts on what success would look like.
Rebecca Endean
One of the key aims of the digital catapult is to help small and medium-size businesses especially have a platform with that. There is a whole range of assets across the research base, including Hartree and various universities, actively engaging with business to try to help them use their data and share the assets they have available to make it easier for them to do that.
Chair
Baroness Shields, one of the big issues running as a thread through the inquiry is what has commonly been called the trust deficit. We heard evidence from the digital catapult that 60% of consumers are uncomfortable sharing their data; 80% of consumers think that organisations gather data simply for economic gain; and the ICO says two thirds of people feel they have lost control of their data. It is your job to run internet safety for the Government, so what do you think we should be doing to try to address the issue of trust, because it seems to be the key barrier in realising some of the undeniable opportunities that are there?
Baroness Shields
You are right to point out the big questions that are generated when everything we do in life is catalogued from devices that are literally becoming an appendage to our lives and minds. It is the business model of many companies to take that data and understand, almost down to your intentions, what you want to do and serve you up the best possible solution around that. With that come lots of challenges. Before we go further into that, I want to point out a study that we did last year called Tech Nation, just to illustrate for the purposes of the Committee the real depth of data science, analytics and machine learning that comes about when you make use of data. The fact is that the UK has 16 centres of excellence across the country. We hear a lot about the northern powerhouse, but we literally have 16 centres in which we have expertise in these areas, in general associated with some of our great universities, but also as a result of working with the digital catapult, Tech City and the organic growth of those businesses supported by Government. Tech City is a low-cost initiative by Government. It has a budget of £2 million. It does not sound much, but the impact of that initiative, putting a spotlight on technology in this country, has led to enormous investment and innovation around this area. We have those 16 thriving centres. On data security, there is a chasm in terms of what people feel about trusting data. If you talk to teenagers, they do not care; they have given up privacy and decided that they are happy to share absolutely everything in their lives and have it catalogued. We have some responsibility to look out for their interests, especially in terms of their rights online, making sure that their rights are served and that there are different rules for young people, because they are not necessarily always making the grown-up decisions we make. On the internet we treat everybody the same. It does not matter if you are seven years old, 25 or 65. Everyone has the same applications and services. That is a real issue.
Chair
The week before last, the Committee went to Oxford as part of Parliament Week. We held a session on big data with 90 sixth-formers who, quite frankly, asked better questions than our Committee did of our experts. One of the questions that arose in terms of the opportunities of big data, but also privacy questions, was where the creepy line between opportunity and privacy is. Where is the creepy line?
Baroness Shields
Society and culture determine where the creepy line is. Tim Cook had a quote. In the summer he had a bit of a rant about technology companies in Silicon Valley—Google, Facebook and companies like that—who literally data-mine every single action we make on the internet. It was quite interesting to hear his perspective. He said that hoovering up all this is far more intrusive than anything Government do, but people do not really see it that way; they do not worry about a commercial company having access to all their data and information, whereas when you talk about bulk collection of data for surveillance purposes everyone is up in arms. He said that on a daily basis we give up much more information than we realise. I think the creepy line is moving further and further out; we accept more and more of our privacy being open.
Chair
Should we be leaving that to society, or should Government be taking action, legislative or otherwise?
Baroness Shields
It is hard to say. It depends. If it is a matter of national security, of course it has to be a Government issue; when it relates to protecting your financial security, absolutely. When it comes to personal issues and what you want to share, it is a little more difficult. It is really up to the individual. With the terms and conditions on every product we use, we always say “Accept.” No one reads what it is being used for. We have been conditioned to say, “Yeah, yeah, thanks; I’d like to use the application,” but the reality is that if you look deep inside it there is probably not a single person in this room who would be comfortable with what is being given up in that case.
Jim Dowd
First, I apologise to you, Chair, and the panel, because I will have to leave shortly. I have a brief question for Baroness Shields. The attitude towards data, even devices, is very much generational, in so far as the younger generation are much more comfortable; older people tend not to be. Is that a social development, in so far as they have been used to it from a younger age, or do you think that as they become older their attitude will become more defensive?
Baroness Shields
I am not sure that the train has not left the station. They are much more accepting of this as a new reality, because you get a lot from giving up data and information; you get a better experience. If you are shopping over the weekend on Black Friday or Cyber Monday and decide not to buy something, you will notice that on the next site you go to that particular product will show up and remind you that maybe you want to buy it. In some sense it is good; it reminds you that the item in the cart you left at the last site you visited is still available if you want to purchase it. In some sense, that gives you an illustration of how much is out there, and that you get a better service. In general, companies can use that data to deliver better services on a daily basis, whether it is train times or something else. We give up a bit but we get a lot as well.
Jim Dowd
You think people will become more relaxed over time.
Baroness Shields
I think so. There are certain things that you definitely do not want to share, but often my worry in the work I do is about trying to educate young people as to what they are giving up and making sure they are comfortable with that decision and that they are old enough to make it.
Carol Monaghan
Some of the companies and organisations that have been referred to—the digital catapult and TechUK—have already spoken to us in previous sessions of the inquiry. These are organisations whose remit is to promote business and entrepreneurship in the field of big data. What is the extent of the Government’s role in promoting entrepreneurship in big data?
Mr Vaizey
We mentioned the catapult. The catapult is looking at all sorts of applications of data. The digital catapult is particularly focused on small businesses. It encourages small businesses and is designed to be a space for them to try out new business models and have the resources available to do that, which only a big company might have. That is the first issue. There are big companies as well. Rebecca referred earlier to the collaboration we are doing with IBM in terms of their health technology. They can work with the big data projects that the Government are funding. There are hundreds of millions of pounds going into these projects, and even big companies could not necessarily match that kind of research, so it is important that we collaborate with business. Given the conversation we have just had about the risks out there, my creepy line would be the difference between identifying an individual as opposed to anonymous data. For me, the big opportunity with big data is huge datasets that allow applications to be put in place that will make a positive difference to citizens’ lives. It is linked to the internet of things and machine-to-machine technology. If you can monitor traffic and link that dataset with air quality, you can provide a better experience for people living in a city, but nobody wants traffic monitoring data to be mined somehow to track where you have been in your car. That is the kind of thing we have to guard against, but it is important for us to work with business. Pure science research is very important, and where business comes in is in the real-world applications.
Carol Monaghan
Do the Government have a specific role in promoting business around big data? Are there boundaries regarding what the Government should and should not be doing?
Mr Vaizey
Rebecca may want to come in with specific examples of how we work with business, but the Government have a threefold role. One is to talk about big data. That may sound a bit facile, but, as Joanna was saying in respect of Tech City, the issues that Government say are important to business do gain traction and attention. Organisations like the catapult have a role in getting out to businesses and showing them how they can use big data. For example, in my own constituency I have the satellite applications catapult. I always dine out on the example they gave me when I first visited them; they engage with supermarkets because supermarkets use satellite imagery to monitor their car parks. The catapults can go out to business, evangelise and say, “You don’t realise how you could be using big data to improve your business performance.” Related to that, Government have to get further down the chain in providing the skillset: for example, in terms of some of the changes we have made to the school curriculum and also to further and higher education. We need to make sure that businesses have available to them people with the right kind of skills to take advantage of this revolution. I do not know whether Rebecca has any examples.
Rebecca Endean
For small businesses, particularly in relation to catapults, there is a clear market failure about helping them get at the technology and taking products to market. We have fantastic assets in the science and research base across all our universities, and fostering really good collaboration between universities and the use of some of these assets, which are basically there only on a national, or even international, scale—for example, Hartree super computing facilities—is clearly a role for Government.
Baroness Shields
To add to what Rebecca and Ed were saying, the Government have great convening power. What we have been able to do in these 16 centres of excellence across the country is bring universities, companies and start-ups together. I was at the Imperial College data science lab, which we launched just last month. You have small businesses engaged in that. At Imperial West there are 60 or 70 small start-ups. All of them are working together with big companies on data visualisations. I saw a Bitcoin visualisation that showed malware interjecting illicit transactions into the universe of Bitcoining. Small companies were reviewing that data and determining what they could do with that and how they could improve their Bitcoin businesses. You see all of this coming together. At Edinburgh University there is collaboration with a lot of e-commerce companies and start-ups. They are all coming together with the university. That is really important. You have the Government, universities, the Alan Turing Institute and the Open Data Institute. During the floods two years ago, the Open Data Institute and the Government Digital Service hosted hackathons with universities and start-ups. People over the course of a weekend created real applications that helped people in need, whether it was moving livestock to higher ground or food and supplies to the right places. All that comes together in the collaboration. We are quite good at that in the UK—[Interruption.]—in spite of this interference we have.
Chair
We have moved on from the creepy line to the creepy sound of the wind outside.
Graham Stringer
The care.data initiative was delayed. Can you tell us what lessons have been learned from that, and how the scheme that is going to be implemented by 2018 will differ from what went before?
Peter Knight
Thank you for inviting me to give evidence. It is for NHS England to comment on the details of lessons learned, because they are the people operating and running the project. What I can do is give you an update on where it is at this point in time. It is currently in what is called a pathfinder phase. There are four clinical commissioning groups across the country looking to pilot what the material and the consultation around it will be used for. I guess the lesson learned is to make sure you get your material right with the population before you go forward. The second thing is that Dame Fiona Caldicott is doing a review of both the consent question and the question around what cyber-security standards we need to operate across the healthcare system. Dame Fiona is taking oversight of what care.data does before it starts to extract data.[1] Until she is satisfied that those two things have been completed, care.data will not be taking data out of the systems at the moment. NHS England will have more detail on that, and certainly I will get them to write to you about the lessons learned.
Valerie Vaz
Minister, I understand that you have to leave early. It would have been quite nice to talk to you about lots of different things.
Mr Vaizey
I am happy to stay. Nothing is more important than this Committee. The all-party writers group will have to wait.
Valerie Vaz
Could you tell us a bit about the digital transformation plan and how we are getting on with it?
Mr Vaizey
The digital transformation plan was announced in the summer Budget. The plan is to produce a report, I hope by the end of January—don’t hold me to that—which will look at drawing together all the digital initiatives the Government are currently undertaking. It will look at all the landscape you are covering in terms of organisations like the Alan Turing Institute, the Open Data Institute and others, as well as Innovate UK, but it will also look at each area where Government operate, whether it is schools, transport or the Home Office, and how we can start to embed digital. The work is being undertaken by officials in the digital economy unit in my Department, but it is important that we work very closely with the Cabinet Office and the Government Digital Service. For me, the great prize of the plan will be essentially embedding digital and fundamentally changing the way the Government do business, but also the kind of experience citizens have in a whole range of different areas.
Valerie Vaz
Will you be looking at privacy issues and regulation as well?
Mr Vaizey
Privacy will form an important part of that. What is clear from all the questions so far in this evidence session is that privacy and security sit right at the heart of everything we do, because it is not just keeping data anonymous; it is also keeping data secure and not allowing it to leak. Cyber-security and keeping data secure is a very important aspect.
Valerie Vaz
Will that work look at the skill shortage in data analysis?
Mr Vaizey
It will certainly look at the skill shortage. We are investing a lot in skills. It is important to say that every developed country is facing a skill shortage. The US is facing one, even though we laud their leadership in technology; other European countries are as well. We are fundamentally changing not just the school curriculum but also further education and university curriculums, embedding a much closer relationship with business; you need that relationship with technology companies because technology changes so quickly. But there is no magic bullet. It will take time. We are working very hard to address some of these issues.
Valerie Vaz
You are straddling two Departments. Which Department has the responsibility for it?
Mr Vaizey
DCMS has responsibility.
Matt Warman
I want to talk a little about how EU data protection regulation meshes with UK data protection regulation. Where are we at the moment? This has been going on for ever.
Mr Vaizey
First, can I note this seminal moment in my parliamentary career? This is the first conversation I have had with Mr Warman that is not about broadband. That is an important point to mark. I think we should return for the anniversary next year. He has led from the front on that. On data protection, we are expecting to complete the trialogues by the end of the year. An important point within your question, Mr Warman, is that DCMS has taken over responsibility for data protection, which from my perspective was a great prize. I do not want to be accused of empire building, but the key for me was to link the DCMS digital technology agenda with data protection because they are, as it were, both sides of the same coin. The trialogues will end this year, and we expect first reading to commence in 2016. It has been a long and very tortuous issue. One of the ways we have made progress is by parking some of the more contentious issues.
Matt Warman
Does that mean we have not signed up to the draft as it is and we are still negotiating on bits and pieces?
Mr Vaizey
We are still negotiating on bits and pieces, but the trialogue is also something we have to take account of. The European Parliament will have strong views as well and we have to steer a sensible middle course.
Matt Warman
What are the sticking points?
Mr Vaizey
The general principles are about the level of burden on business. We do not want to place too many onerous reporting requirements on business. We want to make sure we get that balance absolutely right.
Matt Warman
Do you think our current safeguards are sufficient, regardless of the European regulations?
Mr Vaizey
We can live with them as they are. We do not want the data protection negotiations to drag on for too long. They have already dragged on too long. At the same time, we also want the best result. I do not think it would be sensible to have any kind of interim measures between our current regulations and the future regulation because that would be confusing for business. We want to get on and get the regulation approved and then deal with the equally important aspect of working with business on implementation. It is very important, even when a regulation is approved, that you do not simply sit back and say, “There you are. Get on with it.” You have to have a conversation and dialogue with business about the best way of implementing the regulation.
Matt Warman
There is still the issue of who owns an individual’s data. We trialled that here with midata, and the principle is in the European regulations. How did those trials go, and where do you anticipate us ending up?
Mr Vaizey
We support the principle of getting companies to release useful data. I think we have made a lot of progress with it. We have made some progress with the personal current account market; we have also made some progress with the energy market. Rebecca is nodding her head, so I have obviously read my brief properly. Do you want to add anything?
Rebecca Endean
No, I do not. It’s a perfect reading of the brief.
Mr Vaizey
Do you want to say anything else about the data protection regulations?
Rebecca Endean
The main thing is to make sure it is proportionate and risk-based. In particular, as you know, on the research and science side we are very conscious of the need to protect personal, individual, social and medical data, but at the same time they have massive potential to improve things, like stratified medicine. Peter, I do not know whether you want to say a quick word about how we want to see the EU data protection regulation move on that.
Peter Knight
From a health research point of view, it is absolutely vital that we get a balanced, proportionate regulation, particularly when we are talking about the reuse of data. That is an area where there have been lots of different views across the European sector. It is hugely important that reuse is balanced with privacy and the rights of an individual. Research value is huge. In medical research, we learn a huge amount from the data we collect and use for research in that space. It would put Europe in a difficult position if the secondary uses part was put into a difficult position. Colleagues in DCMS have been fantastic at taking the line forward very clearly and holding the ground on where the line should be. Certainly, the medical research fraternity have lobbied very strongly in this area.
Matt Warman
You touched on the issue of malicious re-identification of individuals. I guess we will have the opportunity to insert that when we put European regulations into UK law. Is there a plan to do that, or is it not an issue from your point of view?
Peter Knight
The key point is that we have more than one area of law to rely on in the reuse of data. We always tend to go for the privacy positioning one, but actually we have contract law and employment law as well, where malicious use of data is occurring. We need to make sure that we use the law appropriately across all three of those different areas, and use the appropriate law for the intent.
Chris Green
Minister, the Data Protection Act provides for civil rather than criminal penalties. Do you think there is a case for introducing criminal sanctions for data protection breaches?
Mr Vaizey
To pick up what Peter said just now, there are criminal penalties under section 55 of the Data Protection Act for the active obtaining of data unlawfully with intent to disclose it. As I understand it, the regulations will give member states an option to introduce criminal penalties where civil penalties currently exist. I have to say in all honesty that I have an open mind about that. To turn something into a criminal offence is always a big step, so we would need to look very carefully at it. As far as I am aware, unless others on the panel want to correct me, we have not come to a conclusion within government on whether or not that would be an appropriate step. We want to get the regulations approved and then return to it.
Chris Green
Are there any particular reasons you would highlight for not doing so?
Mr Vaizey
The fact that it is a criminal offence maliciously to access data with the intent, effectively, to misuse that data covers what perhaps the ordinary person in the street would regard as a criminal act. I would not want criminal legislation inadvertently to catch people who have been negligent, however much they might be condemned for their negligent behaviour in allowing your data to become available. We would have to think very hard, if we were to introduce criminal penalties, about what kind of behaviour they would catch. I stress that I am not ruling out introducing criminal penalties; I am just saying there should be a serious threshold to cross before we decide to go down that route.
Chair
We had evidence from the Information Commissioner on this issue. He has called for a criminal offence to be introduced. He feels that the penalties are not sufficient. His view is that section 55 would not hold up for cases of re-identification where big datasets had been anonymised but then re-identified through jigsaw techniques, or something like that. Is any work going on at the moment in DCMS or BIS properly to test the legislation against the very fast progress going on with data at the moment?
Mr Vaizey
I will write to the Committee on that point. We have recently taken over data protection policy within government, which is a very good thing, as I said earlier. I am not aware if the Ministry of Justice, when they had that portfolio, were undertaking that work. It is important to listen to what the Information Commissioner has to say. The Information Commissioner levies pretty hefty fines on companies, but I am all too well aware that you can spin out the civil process and potentially get away with not paying a fine if your company goes bust and so on. If the Information Commissioner is saying that we should take a serious look at introducing a criminal penalty and if he can make a convincing case that section 55 does not cover what is in effect—echoing your point, Madam Chairman—a new technique because of the rapid way technology can change, we would be very sympathetic to what he has to say.
Chair
I heard what you said about coming to a balanced and proportionate conclusion on the EU data directive. Is there a Government position on explicit consent versus unambiguous consent? There are different European positions on that, as I understand it.
Mr Vaizey
We are concerned about the meaning of explicit consent, and what kind of hoops business would have to go through to introduce it, as opposed to deemed consent or unambiguous consent. We have a similar issue with the e-privacy directive and, in common parlance, how many clicks you would have to go through to show that you have given consent. The current position, which we will all be familiar with when we use websites, is suitable, relatively friction-free and gives you the chance to give consent. We would want to be very clear as to what the Commission meant in practical terms by explicit consent.
Derek Thomas
One of the concerns about big data is that computers may well decide about human life rather than a human being. We might decide that’s not such a bad thing, but when decisions affect human lives—for example, computer-driven credit assessments—are we concerned about that? Is that a problem? Should Government intervene in this area?
Mr Vaizey
There is a philosophical element to your question, Mr Thomas. We know that in many areas of our lives we come across the terrible phase, “Computer says no,” car insurance being an extremely good example. All of us in this room know we are brilliant drivers, but given our demographic or perhaps our children’s demographic, the insurance company, or more accurately the computer, thinks something different. There is a series of different regulators across business who I think should play a role in this. That draws out a very important point: big data and new technology are for everyone and no one should assume it is somebody else’s job. How companies use an algorithm to assess whether you are eligible for insurance and so on should be as much a matter for the insurance regulator to discuss with that business sector as it should be for, say, the ICO or Government Departments charged with the whole big data agenda. From my perspective, my agenda as a Minister working with Lady Shields is to drive research and take‑up of big data to help the UK economy and UK business, mindful of the privacy and security concerns that exist with this kind of technology, but I would take how business uses big data in terms of interacting with its customers almost on a case-by-case basis.
Derek Thomas
Would you say that the upcoming EU regulation provides better protection?
Mr Vaizey
Yes. By definition, I think it will provide better protection and, hopefully, a uniform system of data regulation across the European Union, which will be a very big prize. It is a long overdue update to take into account the fact that companies can now use much more digital data.
Chair
You have been very patient. I am very grateful to you for taking the time out. As a final question, we heard on the previous panel from Hetan Shah, head of the Royal Statistical Society. They have given us evidence that, while the Government are now better at sharing data between Departments, they are less good at sharing it externally. The key recommendation they made in this inquiry is that the Government should open up statistical data-to-data bodies like the Royal Statistical Society and the ONS. They pointed to Canada, which has opened up public data as well as private data, and they believe this is a very good way to develop policy and encourage good use of big data. What is the ministerial response to that proposal?
Mr Vaizey
I am surprised. I would have thought it was almost the other way round. We have made 20,000 publicly held datasets openly available for reuse. I think we came top of the 83 countries surveyed by the World Wide Web Consortium. I think the problem is Government Departments sharing data between Government Departments. We have set up the Government data taskforce with the chief scientist and others to try to get Government Departments to take big data seriously, to see the opportunity and also to share it. Mindful of the ethical concerns Peter talked about earlier surrounding things like care.data, this provides massive opportunities. I think we need to look at potential future legislation to allow that sharing to be made easier between Government Departments. Rebecca, you may have a view. I hope it is not different from mine.
Rebecca Endean
It is not. When we talk about physical data and stuff like data-backed themes, we have been very good. Where there is an issue, which is not surprising, it is the personal data we hold on individuals in areas of health, higher education and education. Where I used to work—the Ministry of Justice—we hold great datasets in government, but we have to hold them very securely because they include very sensitive data. We are trying to explore ways of making that data available to academe in a way that is safe and in accordance with the law, and also bears in mind the important ethical and privacy issues academics take very seriously. There is also quite a lot of work on growing metadata so you can work out what the data is telling you. A lot of people across government, including the Cabinet Secretary, are very keen to make this happen and progress is being made. Some of the data protection and privacy issues related to these datasets are very important and we need to treat them very carefully.
Chair
Clearly, privacy issues are important; we are spending a lot of time talking about them, but are the Government or different Departments too risk averse? Are some more risk averse than others? Is there a cultural problem that needs to be overcome, or is it merely a technical or legislative one?
Rebecca Endean
There is varying progress. If I could put in a plea for my old Department, the Ministry of Justice, they developed a very interesting and novel way of helping charities work out who is and is not reoffending, by allowing charities to send their data to the Ministry of Justice. The Ministry of Justice did the matching and analysis and sent back the results. That was hugely successful. There are various data labs like that growing up across government. In addition, there are ways of working with the ESRC and others to try to get big datasets out. There is always more you can do. Sorry, this is my favourite subject so I’m ranting. I will pass over to Peter so he can have a word.
Peter Knight
There is also an issue about compute over, as opposed to taking data away; you can run your algorithm and get your result back and use it in situ. Health is working a lot to look at that area, because of the privacy issues and the confidential personal information held there. The concept of being able to compute over and get a result out gives you a balance between safety and utility.
Mr Vaizey
The answer is yes, but you can understand why.
Baroness Shields
I am curious about whether that comment related to commercial opportunities or to better services for citizens, because they are different. If it is commercial data we are doing really well, but when it comes to utilising personal data it is very difficult to make sure that it is anonymous and you are protecting privacy. There are so many new techniques to use machine learning to repurpose the data and construct who that individual was, backwards from the point of it being anonymous. We have to be really careful with that. I would err on the side of a little more caution when it comes to personal data anonymised for commercial purposes. I would be more than happy to suggest that we give more to other Departments so that people can get a better service from Government, but we have to tread carefully. We have pushed the boundaries more than any other country, and that is really positive and commendable. I am sure there is more we can do, but if it is for commercial purposes we are doing pretty well.
Chair
We have come to the end of our session. Thank you for the time you have taken, and for your patience in waiting for us while we had two votes on the Immigration Bill—shocking—and I thank members for their brief questions during the end of this session. That brings us to the end of the big data dilemma inquiry. The conclusion we have probably come to in this session is that we must not be so occupied with managing the risks of big data that we manage out the opportunities, but clearly I cannot predetermine what the Committee will conclude in our report. We may come back to you with further questions as we prepare our report. Thank you for your time. Oral evidence: The big data dilemma, HC 468 24 [1] The witness later clarified that, Dame Fiona is setting the criteria for NHS England on Information Governance for care.data to achieve before it starts to extract data.