Cabinet Office Minister Matt Hancock announces new framework based on six good practice principles
User benefits, respect for privacy and security should be key features in the growing use of data for making policy and operational decisions, according to a new framework published by the Cabinet Office.
Its Data Science Ethical Framework outline six principles aimed at helping government organisations and others to deal with the legal aspects of thinking through some of the issues that, while sitting outside the law, could cause problems if badly handled.
The need for a framework has been growing as government has taken an increasing interest in data science along with its effort to build a national data infrastructure. The Government Digital Service has been leading the effort to promote its potential, and recently teamed up with the Office for National Statistics (ONS) and the Government Office of Science in forming the Government Data Science Partnership.
There have also been moves to set up a Data Science Campus under ONS and create a Council of Data Science Ethics.
The six principles in the new framework include starting any project with a clear view of user and public benefit, and using data and tools that create the minimum level of intrusion. They also involve the importance of keeping data secure, being alert to shifting public perceptions around how data is used and levels of consent, and being as open and accountable as possible about tools, data and algorithms.
The sixth is to create robust data science models, thinking through any possible unintended consequences and flagging if algorithms use protected characteristics, such as ethnicity, in making decisions.
The framework also includes a checklist for complying with each principle, and a six-point questionnaire that can act as a privacy impact assessment.
Potential and anxieties
In a speech announcing the publication of the framework, Cabinet Office Minister Matt Hancock said there is great potential in the use of data but that transformation is disruptive and causes anxieties.
“We want to unlock the progressive power of data science to improve lives,” he said, speaking at an event staged by innovation charity Nesta. “And we want people in government to feel confident using new techniques.
“This means setting out clear guidance that brings together the relevant laws and best practice, and gives data scientists and their teams robust principles to work with. It is all about encouraging new and innovative ways to better solve problems and deliver.
“So today we are launching our new Data Science Ethical Framework, setting out in one place our framework for using data. It will help people using data to ask the right questions and take appropriate steps.”
He added that is a first version and will be subject to iterations to reflect changes in technology and the way data is used.
Image from GOV.UK, Open Government Licence v3.0