Government must help citizens challenge algorithm decisions, MPs conclude

Select committee report calls for algorithms to be added to a ministerial brief, and urges departments to publicly declare where and how they use them


Photo: Pixabay

By Sam Trendall

24 May 2018

MPs have urged the government to consider introducing a legally enforceable “right to explanation” that allows citizens to find out how machine-learning programmes reach decisions affecting them – and potentially challenge their results.

A House of Commons select committee has also urged the civil service to publicise where and how it is using algorithms, and for a “ministerial champion” to be appointed to oversee and manage their use.

The Science and Technology Committee has published a report on the use of algorithms in decision making which recommends that “the government should play its part in the algorithms revolution in two ways”.


RELATED CONTENT


The first of these is continuing to make public-sector datasets available for use by developers of algorithms and big-data programs. 

The second role for identified by the committee is that “the government should produce, publish, and maintain a list of where algorithms with significant impacts are being used within central government, along with projects underway or planned for public-service algorithms”.

This, the MPs said, will provide greater transparency, and will make it easier for private-sector firms to work with the government.

A named government minister should also be given responsibility for overseeing government’s use of algorithms and ensuring a joined-up approach from Whitehall departments.

The report said: “The government should identify a ministerial champion to provide government-wide oversight of such algorithms, where they are used by the public sector, and to co-ordinate departments’ approaches to the development and deployment of algorithms and partnerships with the private sector.”

"Algorithms can make flawed decisions which may disproportionately affect some people and groups"
Norman Lamb, committee chair

The report also recommends that the Crown Commercial Service should undertake a review – which could be conducted by the Alan Turing Institute – with the aim of establishing an appropriate procurement model for algorithms.

The government’s soon-to-launch Centre for Data Ethics and Innovation must also play a key role in ensuring that the use of algorithms is subject to “accountability and transparency”, the committee concluded.

The centre’s first job should be to assess the various tools currently available for auditing algorithms, MPs said. It should then report back to government “on which to prioritise and on how they should be embedded in the private sector, as well as in government bodies that share their data with private-sector developers”.

Another task for the data ethics body – in tandem with the Information Commissioner’s Office – should be to consider how the law might be adapted to ensure citizens have the ability to interrogate and challenge the decisions made by algorithms and, if necessary, seek recompense for the effects caused.

The provisions for a ‘right to explanation’ made by the EU General Data Protection Regulation, which comes into effect on Friday, are not legally binding, MPs said. 

“The right to explanation is a key part of achieving accountability,” the report said. “We note that the government has not gone beyond the GDPR’s non-binding provisions and that individuals are not currently able to formally challenge the results of all algorithm decisions or, where appropriate, to seek redress for the impacts of such decisions. The scope for such safeguards should be considered by the Centre for Data Ethics and Innovation and the ICO in the review of the operation of the GDPR that we advocate.

The report recognises that sharing a comprehensible explanation of the workings of algorithms could create “commercial or personal-data confidentiality issues” for their developers. But the use of so-called data trusts – which would set out clear rules and standards for how sensitive data could be safely shared between parties – could enable such issues to be overcome, the MPs said.

A data trusts model was proposed in a government-commissioned review of the AI industry published late last year.

“The government and the Centre for Data Ethics and Innovation should explore with the industries involved the scope for using the proposed data-trust model to make that data available in suitably de-sensitised format,” the report said. “While we acknowledge the practical difficulties with sharing an explanation in an understandable form, the government’s default position should be that explanations of the way algorithms work should be published when the algorithms in question affect the rights and liberties of individuals.”

"Individuals are not currently able to formally challenge the results of all algorithm decisions or, where appropriate, to seek redress for the impacts of such decisions. The scope for such safeguards should be considered."
Algorithms in decision making report

Other recommendations made by the report include the undertaking of more government-backed research into the potential benefits and risks of algorithms, as well as an “immediate review of the lessons of the Cambridge Analytica case” to be conducted by the ICO and the Centre for Data Ethics and Innovation. The two bodies should also review the existing regulatory environment and help guide regulators in key sectors.

Committee chair Norman Lamb said: “Algorithms present the government with a huge opportunity to improve public services and outcomes, particularly in the NHS. They also provide commercial opportunities to the private sector in industries such as insurance, banking, and advertising. But they can also make flawed decisions which may disproportionately affect some people and groups.

He added: “The government must urgently produce a model that demonstrates how public data can be responsibly used by the private sector, to benefit public services such as the NHS. Only then will we benefit from the enormous value of our health data. Deals are already being struck without the required partnership models we need."

Read the most recent articles written by Sam Trendall - MoD to investigate potential for AI to improve productivity

Share this page