Data Protection Bill [HL]

Lord Mitchell Excerpts

Report: 3rd sitting Hansard: House of Lords
Wednesday 10th January 2018 Lords Chamber

Moved by

[Lord Mitchell]

107B: Schedule 13, page 186, line 23, at end insert—
“(j) maintain a register of publicly controlled personal data of national significance;(k) prepare a code of practice which contains practical guidance in relation to personal data of national significance.(2) For the purposes of sub-sub-paragraphs (j) and (k) of paragraph (1), personal data controlled by public bodies is data of national significance if, in the opinion of the Commissioner, —(a) the data furthers collective economic, social or environmental well-being,(b) the data has the potential to further collective economic, social or environmental well-being in future, and(c) financial benefit may be derived from processing the data or the development of associated software.”

[Lord Mitchell (CB)]

My Lords, I will also speak to Amendment 108. The points I am addressing were glossed over in Committee, and I now wish to expand on this important issue.

Data is the new oil. This has been said many times in your Lordships’ House, but as each day passes it becomes more true. Without stretching the analogy too far, in our country big data is about to become the 21st-century equivalent of North Sea oil. Because big data has such value, it will come as no surprise to see big tech companies swarming all over it. They have to because it is their lifeline. Many of our public bodies, particularly the NHS, are custodians of massive amounts of data, which big tech is eager to get its hands on. But we as legislators who act for the public good also have a responsibility to ensure that the public are protected and that, simply put, our treasure is not taken from us without clear authority or appropriate recompense. The data the public bodies hold belongs to us all. It is ours—our communal property—and we must tread carefully.

I will make one point as strongly as I can. I am a product of the data revolution; I have been professionally involved in the digital industry for over 50 years. For 40 of those I was an IT serial entrepreneur. This industry has been good to me; I fully understand that the tech sector needs light regulation. I know that at its best the digital revolution is a force for good but, equally, I know the dangers it poses, so I am trying to be cautious in what I propose. We stand at a crossroads. Computing power has reached astronomical capabilities, software is increasingly complex and artificial intelligence is now making dramatic inroads. Plus, we see the exponential availability of digital data. All these have contributed to the creation and brilliance of algorithms. The one thing we know for certain is that these exciting developments will keep on growing at exponential rates. In medicine, for example, new tools are being developed that are already enhancing diagnostic and treatment capabilities that could benefit all manner of healthcare, in particular our ageing population.

I welcome these developments, as I am sure we all do, many of which have come from our own private sector, and we should rejoice at this example of British expertise. However, at the same time we need to strike a balance between the ambitions of 21st century businesses and the responsibility of government to steward assets and resources of national significance so that the proceeds of technological developments benefit us all. My two amendments seek to codify how valuable, publicly controlled personal data is shared with big tech companies, and to ensure that financial returns, combined with wider social, economic and environmental benefits, are optimised.

I can best demonstrate the scale of this issue if I refer to the NHS. Ever since its formation in 1948—maybe they were kept even before that—the NHS has kept records of tens of millions of patients, literally from cradle to grave. These records are either in written form, or increasingly in digital format, but the magnitude of the collected data is huge. Very few countries can match the length and depth of the health records that the NHS is trusted to retain on behalf of the general public. Such data is called longitudinal data and, when it is bundled together, has great commercial value.

At Second Reading I gave the example of a company called DeepMind, which is a British subsidiary of Google. I visited DeepMind, which is an impressive organisation based here in London. It has purchased access to millions of anonymised data records from institutions such as the Royal Free and Moorfields Eye Hospital. It does not buy this data outright—it does not have to. It simply buys access. Such access enables it and companies like it to use very powerful computers and very sophisticated software to process millions of records with the help of artificial intelligence and machine learning.

This synthesising of data using AI capabilities is designed to produce algorithms, and it is these algorithms that become the product that companies such as DeepMind are able to monetise. They do this by selling the algorithms and their consulting services to the likes of pharmaceutical companies and healthcare providers and even back to the NHS itself. It is a global business and very profitable. At the Royal Free, these algorithms are being used to detect the early onset of kidney disease. At Moorfields Eye Hospital, also here in London, spectacular advances have occurred in similarly detecting potential optical problems.

This is data processing used for the benefit and enhancement of all mankind and we should welcome it. However, I am concerned that this precious and unique data is being offered to big tech companies by our public bodies in the absence of clear and consistent guidelines and without asking how best to obtain value for money in the broadest sense of the term.

Having dealt with big tech companies for most of my life, I know that they are staffed with exceptionally clever people and are no slouches at driving hard bargains. Unlike our NHS, they are not consumed with the day-to-day preoccupation of trying to balance their current budgets; with hundreds of billions of dollars in the bank, they can afford to play the long game, and it is easy to see who holds the aces in any negotiation. Put simply, I wish to protect our public bodies and ensure that we do not give away our inheritance. That is why we need to codify how we will obtain value for money from the sharing of data of national significance with the private sector.

My proposal is not just for the NHS and it is not just for now. All public bodies need protection and guidelines today and well into the future. That is why I have introduced my amendments. In Amendment 107B I seek, first, to require the Information Commissioner to maintain a register of publicly controlled personal data of national significance and, secondly, to prepare a code of practice containing practical guidance in relation to personal data of national significance. These are defined in subsection (2). In Amendment 108 I have set out the requirements of the code on personal data of national significance.

[Lord Clement-Jones]

My Lords, I want briefly to express sympathy with the noble Lord, Lord Mitchell. I share many of his concerns but essentially I think that we should look on the most optimistic side. I hope that he is also really describing the opportunities that can be made available with this kind of data, provided that it is accessible in the way described. I know that the noble Lord takes considerable inspiration from Future Care Capital’s report on intelligence-sharing unleashing the potential of health and care data in the UK to transform outcomes. I thought that it was very good and well considered.

The noble Lord has put down a very important marker today but my one caveat is that I am not sure that there is yet a settled view about how to deal with this kind of data. In Committee we talked about data trusts. In her AI review, Dame Wendy Hall also talked about data trusts. I know that we need to head in a direction that gives us much more assurance about the use of the data in the way that the noble Lord, Lord Mitchell, has described, but I am not sure we have quite reached a consensus around these things to come to the decision that this is the best possible model.

--- Later in debate ---

[Lord Ashton of Hyde]

My Lords, I am grateful to the noble Lord, Lord Mitchell, for taking the time to come and see me to explain these amendments. We had an interesting conversation and I learned a lot—although clearly I did not convince him that they should not be put forward. I am grateful also to the noble Lords, Lord Clement-Jones and Lord Stevenson, who said, I think, that there may be more work to do on this—I agree—and that possibly this is not the right time to discuss these issues because they are broader than the amendment. Notwithstanding that, I completely understand the issues that the noble Lord, Lord Mitchell, has raised, and they are certainly worth thinking about.

These amendments seek to ensure that public authorities—for example, the NHS—are, with the help of the Information Commissioner, fully cognisant of the value of the data that they hold when entering into appropriate data-sharing agreements with third parties. Amendment 107B would also require the Information Commissioner to keep a register of this data of “national significance”. I can see the concerns of the noble Lord, Lord Mitchell. It would seem right that when public authorities are sharing data with third parties, those agreements are entered into with a full understanding of the value of that data. We all agree that we do not want the public sector disadvantaged, but I am not sure that the public sector is being disadvantaged. Before any amendment could be agreed, we would need to establish that there really was a problem.

Opening up public data improves transparency, builds trust and fosters innovation. Making data easily available means that it will be easier for people to make decisions and suggestions about government policies based on detailed information. There are many examples of public transport and mapping apps that make people’s lives easier that are powered by open data. The innovation that this fosters builds world-beating technologies and skills that form the cornerstone of the tech sector in the UK. While protecting the value in our data is important, it cannot be done with a blunt tool, as we need equally to continue our efforts to open up and make best use of government-held data.

In respect of health data, efforts are afoot to find this balance. For example, Sir John Bell proposed in the Life Sciences: Industrial Strategy, published in August last year, that a working group be established to explore a new health technology assessment and commercial framework that would capture the value in algorithms generated using NHS data. This type of body would be more suitable to explore these questions than a code of practice issued by the Information Commissioner, as the noble Lord proposes.

I agree that it is absolutely right that public sector bodies should be aware of the value of the data that they hold. However, value can be extracted in many ways, not solely through monetary means. For example, sharing health data with companies who analyse that data may lead to a deeper understanding of diseases and potentially even to new cures—that is true value. The Information Commissioner could not advise on this.

That sharing, of course, raises ethical issues as well as financial ones and we will debate later the future role and status of the new centre for data ethics and innovation, as the noble Lord, Lord Stevenson, mentioned. This body is under development and I am sure that this House would want to contribute to its development, not least the noble Lord, Lord Clement-Jones, and his Select Committee on Artificial Intelligence.

For those reasons, I am not sure that a code is the right answer. Having heard some of the factors that need to be considered, I hope the noble Lord will not press his amendment.

Perhaps I may offer some further reassurance. If in the future it emerged that a code was the right solution, the Bill allows, at Clause 124, for the Secretary of State to require the Information Commissioner to prepare appropriate codes. If it proves better that the Government should provide guidance, the Secretary of State could offer his own code.

There are technical questions about the wording of the noble Lord’s amendment. I will not go into them at the moment because the issues of principle are more important. However, for the reasons I have given that the code may not be the correct thing at the moment, I invite him to withdraw his amendment.

[Lord Mitchell]

My Lords, I thank all noble Lords for their contributions to this short debate. I also thank the Minister for agreeing to see me prior to the Recess and for his comments today. However, this is an issue of precision—and we need precision on the statute book. All that has been suggested to me, which is that it can be found elsewhere or will be looked at in the future, does not give the definitive answer we require. That is why I would like to test the opinion of the House.

--- Later in debate ---

Moved by

[Lord Mitchell]

108: Before Clause 119, insert the following new Clause—
“Code on personal data of national significance
The Commissioner must prepare a code of practice which contains—(a) best practice guidance in relation to information sharing agreements between publicly funded data controllers and third parties;(b) guidance in relation to the calculation of value for money where publicly funded data controllers enter into information sharing agreements with third parties;(c) guidance about securing financial benefits from the sharing of such personal data with third parties for the purposes of processing or developing associated software, and(d) such other guidance as the Commissioner considers appropriate to promote best practice in the sharing and processing of personal data of national significance.”