Using research to shape the future of artificial intelligence
What led you to work on this particular research project?
With AI being adopted in the public sector, we are potentially witnessing a pivotal change. Using AI in government can bring benefits (more accurate decisions, more personalised services) and risks (more surveillance, more discrimination).
Governments need to act both quickly and wisely to shape this opportunity for the best of their societies. In Europe, policymakers are trying to realise this by involving different stakeholders, including academia, to define the governance of and with AI in the public sector.
This is why I welcomed the opportunity to contribute as an expert in the AI Watch, an initiative of the European Commission to monitor and facilitate the implementation of AI in member states.
What is the impact of your research so far?
I have had the opportunity to share research-based inputs by chairing a peer-learning workshop with more than 80 public officials from 23 EU Member States. Also by serving on the Advisory Board for the establishment of the new Master in Artificial Intelligence for Public Services (AI4Gov), whose first students are graduating this year. And, earlier this year, by engaging in a global discussion at the European AI Excellence and Trust in the World event at the Dubai World Expo.
A key ambition is to also shape public policy. Together with researchers and policy officers from across Europe, we published a comprehensive handbook containing 16 actionable recommendations for European public managers at all levels on how to use AI in public services in an effective and human-centric way.
Why is this topic important?
AI techniques are already providing substantial benefits. Machine learning, for instance, can help doctors formulate more successful cancer treatments. Conversational AI chatbots can upscale personalised services to citizens that do not have digital skills.
But there is also a dark side in the potential risks of AI. For example, the indiscriminate use of automatic facial recognition in public places can lead to undue surveillance. Profiling citizens using behavioural data to flag people likely to commit tax fraud can bring about discrimination.
This is not a futuristic scenario, but something that is already happening. AI Watch has already analysed 686 cases, both successful and less successful, of AI use in the public sector in Europe.
To whom is your knowledge relevant?
Obviously, there is an immediate relevance for academic disciplines such as Digital Government, Public Administration, and Information Systems.
But most importantly, understanding benefits and risks of AI is crucial to public managers, both as regulators and as providers of services. To businesses, both as users and as technology vendors. And to citizens, both as recipients of public services, and as co-creators of public policy.
What longer term impact could this have on society?
As we already start seeing, AI in the hands of government can go both fantastically right, or horribly wrong. To get it right, we need to combine two things: fostering experimentation with AI, and adopting an approach to AI that puts the human at the centre.
The former is needed for Europe to stay at the forefront of global innovation competition. The latter is needed to embed in the technology our European values of dignity, fairness, and sustainability.
Read the full report “AI Watch, road to the adoption of Artificial Intelligence by the public sector: A handbook for policymakers, public administrations and relevant stakeholders” here.
To find out more about AI Watch click here.