Report showcases public sector AI use cases

The public sector has trailed the private sector in the adoption of Artificial Intelligence (AI), but governments are making progress in catching up, according to a primer on AI in the public sector published last week by the OECD’s Observatory of Public Sector Innovation in Paris.
Of 50 countries surveyed, the report found that 36 have, or plan, public sector AI strategies that will allow AI to be integrated into policy-making and service design. Innovations identified include Belgium’s use of AI to analyse citizen input on issues such as climate change, and improving the security of Canadian air cargoes.
Belgium’s CitizenLab is a civic technology company that uses machine-learning algorithms to help civil servants easily process thousands of citizen contributions. The platform dashboards can cluster inputs by theme, demographic classification or location. In early 2019, CitizenLab set up a participation platform in response to widespread Belgian protests over lack of action to tackle climate change. Users were invited to submit their ideas, and more than 1,700 were forthcoming over three months. AI was used to pick out the best ones, and CitizenLab is using the findings to develop a report for elected officials with 16 policy recommendations.
Few governments have the resources to scan air cargo records for risk before loading, but Canada has been using AI to address the logistical problem. Transport Canada’s Pre-load Air Cargo Targeting (PACT) team receives nearly one million pre-load air cargo records per year. Each record may include anywhere from ten to 100 fields. The report estimates that a human employee does not have enough time to review even 10% of the records.
While AI cannot yet replace human analysis of cargo, it showed more efficient results in filtering and prioritisation than simple Excel filters. AI can catch anomalies and nuances in a way that Excel cannot. Since the testing phase, the team has produced a dashboard and a first version of a targeting interface for identifying potentially high-risk cargoes. Transport Canada is now working to integrate the approach into its risk assessment process.
OPSI’s Case Study Platform collects examples of public sector innovation. Anyone in the public sector may submit ideas to the platform. It currently houses over 300 cases, about 30 of which include an AI component.