Seite - 572 - in VULNERABLE - The Law, Policy and Ethics of COVID-19

VULNERABLE572 modélisation éthique et les données responsables. Dans une perspec- tive mondiale, il met en lumière des questions particulièrement perti- nentes pour les pays qui étaient déjà en situation de crise humanitaire (insécurité alimentaire, conflits, etc.), en expliquant comment ces questions influent sur la manière dont les modèles épidémiologiques devraient être construits. En outre, ce chapitre se penche sur la vul- nérabilité du point de vue des bénéficiaires de l’aide et des migrants, pour évoquer le type de lignes directrices et de lois qui devraient être prises en compte pour assurer la protection des données et de la vie privée. The COVID-19 pandemic has brought to light vulnerabilities in all domains, while also showing the promise of strong, thoughtful solutions. Artificial intelligence, one of the most impactful research areas of the last decade, is no exception. The COVID-19 crisis is being called a “data-driven pandemic.”1 As countries seek to understand, confine, find cures, develop vaccines, protect the vulnerable, and de- confine the population, the use of artificial intelligence-based solu- tions has been very appealing, and in some cases, disturbing. Fundamentally an analytical tool, what we call artificial intel- ligence in this context, is actually a hybrid of computer science and statistics that allows for rapidly evolving models that learn and adapt over time. In large part due to stronger computational power and massive data storage capacities, a series of powerful techniques have emerged that allow for fast and more accurate predictions in many domains. It should therefore come as no surprise that, at a time where uncertainty and doubt has been universal, computational modelling has gone mainstream. For instance, artificial intelligence models are being used in three medical research fields—testing, treatment, and vaccinations.2 As early as January 2020, BenevolentAI’s algorithms3 were combing 1. Roberto Rocha, “The Data-Driven Pandemic: Information Sharing with COVID- 19 is ‘Unprecedented’”, CBC News (17 March 2020), online: <https://www.cbc.ca/ news/canada/coronavirus-date-information-sharing-1.5500709>. 2. Swati Bhonde, Jayashree Prasad & Madhulika Bhati, “Predictive Analytics to Combat with COVID-19 Using Genome Sequencing” (April 15, 2020), online: SSRN <https://ssrn.com/abstract=3580692>. 3. “Benevolent AI Website” (May 2020), online: Benevolent AI <https://benevolent. ai>.