All Party Parliamentary Group | AI Evidence Meeting
Data Governance in the Post-Brexit Era:
Is the National Data Strategy ambitious enough ?
January 25th 2021
Ross Edwards – IORMA Technology Director and IORMA iLab Director
Lord Clement-Jones and Stephen Metcalfe MP chaired the meeting as usual along with Professor Birgitte Andersen from The Big Innovation Centre
Dr Jeni Tennison, Open Data Institute
Dr Jeni Tennison talked about the UK’s institutional infrastructure for collecting data, the importance of maintaining it and sharing it appropriately. To support the goal outlined by the AI council in making the UK one of the best places to work and live with AI.
Jeni spoke about how data should be viewed as a new form of info-structure similar to electricity or transport and the importance of making sure there is a trusted eco system in place.
She also mentioned public and private institutions should work together regarding data infrastructure and how we could learn from such things as the Information Banks in Japan.
Jeni also pointed out three key objectives:
1. Organisations need to be fit for purpose.
2. That at the moment they might not see themselves as data collectors but must start to recognise this to help their industries.
3. Explore these new forms of best practices around topics such as data governance.
Dr Chris Francis, SAP
Dr Chris Francis talked about how AI was being used and taken up in some sectors but a long way behind in others. Things like pre-filled forms and connecting things like health and insurance information are areas that need to be brought up to speed and that generally services in the government are not joined up.
The UK is in the bottom quadrant of European countries in use of AI in companies.
Chris made the point that AI has to deal with unique situations that need special training.
He noted that 75% to 80 % of UK companies are not using AI and the knock on effect is that they will be less prepared for the future as the training data will not be there or good enough moving on.
He finished by emphasising the need for a clear Data strategy and it needs to be shared.
Professor Edga Whitley, London School of Economics
Professor Edga Whitley is concerned about data sharing especially when it comes to personal data.
He pointed out that the public service delivery path is massively underused and that legal gateways for sharing data are also hardly being used.
He added we should be concerned about the basic competences we have in using AI and that data is not being regarded or seen as an important strategic asset at this moment.
Edga said that currently poor data is having to be worked around to avoid the garbage in garbage out scenario and that data protection laws need to be fit for purpose.
He also feels there needs to be direction from the government rather than consensus as that would be subject to bias from who is doing it.
Del Alibocus, Capita
Del Alibocus is hoping for a positive outcome for a comprehensive strategy similar to the 5G one.
He went on to say he recommends creating a National Data Streaming Service and also recognising it as critical infrastructure and to have the ability to gather the data from data and then by using AI to get the intelligence from that data.
Gathering the data he said is not the issue, rather it is making sense of it, that is the important thing.
Local and central government will then have the tools to have data lead initiatives that allow the tracking of new programs and policies. To help implement this it might be better to concentrate on a couple of sectors to begin with.
Del said we should look at data as a currency and that we must have a data exchange in this country; it is so important that he feels it needs its own Ministry in Government.
Dr Adrian Weller, The Alan Turing Institute
Dr Adrian Weller talked about data and AI having the potential to help society both socially and economically and the importance of the National Data Strategy being very important as the country looks to recover from the pandemic.
He also spoke about the importance of working on reducing unwanted bias in the data being used and a CDI report setting out five key things for data sharing projects;
1. Value.
2. Security.
3. Accountability.
4. Control.
5. Transparency.
Adrian mentioned that there needs to be a balance between sharing data quickly and that of data rights.
He also spoke of The Alan Turing Institute’s response to the National Data Strategy recommending a clear mandate for UK values and principals.