Leveraging AI in the Age of Pandemics. Beating Covid-19, Preventing the Next Crisis

Policy Brief

AI tools are significantly helping governments, the medical community, companies, other organizations and citizens respond to the current pandemic crisis in several different ways.

A recent OECD paper listed the main AI applications currently used: I. understanding the virus and accelerating medical research on drugs and treatments; II. detecting and diagnosing the virus and predicting its evolution; III. assisting in preventing or slowing the virus’ spread through surveillance and contact tracing; IV. responding to the health crisis through personalized information and learning, fighting misinformation; V. monitoring the recovery and improving early warning tools.

I. AI could allow scientists to analyze huge volumes of data in a much shorter time. Some examples of AI use in this area:

• deep learning models can help uncover or predict old and new treatments for Covid-19;

• access to datasets in epidemiology, bioinformatics and molecular modelling is being provided;

• computing power is also being made available by tech companies, research institutions and public-private initiatives;

• innovative approaches including prizes, open-source collaboration and hackathons are actively boosting AI-based research.

II. Algorithms that identify patterns and anomalies can detect and predict the spread of pandemic, while image recognition systems are speeding up medical diagnosis. Some examples of AI use in this area:

• AI-powered early warning systems can help detect epidemiological patterns by mining mainstream news, online content and other information channels in multiple languages to provide early warnings;

• AI tools allow the identification of virus transmission chains and infer epidemiological more rapidly than traditional reporting of health data;

• applied to images and symptom data, AI could help in rapidly diagnosing Covid-19 cases.

III. AI applications are contributing to preventing the virus’ spread.

Some examples of AI use in this area:

• a number of countries have been using population surveillance tools, powered by AI, to monitor Covid-19 cases. For example, South Korea has used geolocation data, surveillance-camera footage and credit card records to trace coronavirus patients. China has assigned a risk level to each person, indicating contagion risk;

• many countries have set up contact tracing apps to rapidly identify possible contagion hotspots and prevent them from spreading elsewhere. In Europe, after sorting out the technical and privacy issues and a trial period, many contact tracing apps are expected to become fully operational in the next weeks;

• semi-autonomous robots and drones have been deployed to respond to immediate needs in hospitals or quarantine facilities;

• wearables and apps may enforce social distancing rules at work or in public spaces.

IV. AI tools can supply better information and fight disinformation. Some examples of AI use in this area:

• virtual assistants and chatbots have been deployed to support healthcare organizations;

• vulnerable and high-risk individuals have been identified and contacted by platforms addressing possible complications from pre-existing conditions and social isolation;

• social networks and search engines have been using AI tools to fight misinformation spreading online (the so-called “infodemic”).

V. AI tools can also be used to monitor the economic crisis and the recovery (e.g. Google’s Community Mobility Reports) and learn from the current crisis the main means for setting up early warning systems for future pandemics. Algorithmic economics – the machine learning applied to economic modelling – could be a new promising frontier for research in economics and the current pandemic is contributing to driving it to make better and faster predictions in such a volatile period.