The Full report can be downloaded Here.

As a disruptive technology, Artificial Intelligence (AI) has been changing all aspects of humanity and society, and is expected to promote the establishment of a “healthy society.” The word “healthy” refers to both public health and people’s physical and psychological health, and the harmonious, healthy and sustainable development of relations between people and nature, society and the environment. Advanced technology should be developed for promoting the establishment and development of a “healthy society.” To achieve this goal, special attention must be given to the potential risks involved in AI development and applications.

In this regard, the Research Centre for Artificial Intelligence Ethics and Safety at the Beijing Academy of Artificial Intelligence, in collaboration with the China-UK Research Centre for AI Ethics and Governance of Institute of Automation of Chinese Academy of Sciences, initiated this survey and study on AI and healthy society, aiming to contribute to the understanding and building of a healthy society driven by AI through research and analysis.

Facial recognition, an important branch and application of AI technology, has shown its great effect in national and public security, daily life and other scenarios. However, it also brings risks and challenges in terms of privacy and security. While during the COVID-19 outbreak, facial recognition and other automated detection technologies have played an important role in epidemic prevention and control, applications related to citizens’ personal information and privacy should be managed in compliance with the laws and regulations.

Definitions in this survey: A “public health crisis” is an event similar to the COVID-19 pandemic. A “normal situation” is a circumstance where no public health crisis happens. A “public space” is a place that is usually open to the general public, such as sidewalks, parks, and government buildings. A “non-public space” refers to blocks and buildings that are privately owned or not open to the general public.

Featured as the first report in Survey Series on Artificial Intelligence and Healthy Society, this report focuses on Facial Recognition and Public Health. The aim of this report is to reflect the focus and thoughts of the general public on this topic, and promote attentions and considerations to AI Ethics and Governance during the development, use, and deployment of facial recognition services for public health.

Key Findings

Based on statistical analysis from the research and other resources, the key findings can be summarized as follows.

  • The respondents generally appreciate the benefits of installing facial recognition in areas with public (health) security concerns. Most of the respondents acknowledge public safety as the primary reason for using facial recognition. The respondents generally agree that those who cover their faces intentionally during a public health crisis should be punished., and most respondents are in favor of the enhanced facial recognition (since the research was conducted during the COVID-19 public health crisis, results might be affected in terms of facial recognition of masked faces and temperature checks).
  • The respondents are concerned about their privacy related to the uses of facial recognition. More than half of the respondents express concern over their privacy in the uses of facial recognition and hope they can understand and control the use of data of their facial features, although the majority of them do not work in industries or research related to facial recognition and only a few are well-informed of the technology. It shows that privacy is still a main concern even in a public health crisis.
  • The COVID-19 public health crisis improved the acceptability of facial recognition among the general public. During a public health crisis, the respondents are more willing to accept facial recognition. They acknowledge the positive role the technology plays, as well as enhanced facial recognition, but they still expect their privacy to be well-protected.
  • The respondents hope the application of facial recognition can be reduced when the pandemic ends. They believe data of facial features collected at this special time should be deleted afterwards, and that unnecessary facial recognition applications eliminated.
  • Some of the respondents are not sure whether they support the reduction of facial recognition applications related to public health crises when the pandemic ends, which shows their concern over the recurrence of such public health crises and similar potential crises. This also gives weight to appreciating the potential of such technology to be prepared for future emergencies. Other respondents remain “neutral” or “strongly disagree” in terms of reducing facial recognition usage when the emergency ends, which highlights the necessity to establish a public health crisis precautionary and defense system driven by technology in order to prepare for any recurrence.
  • The acceptability of facial recognition applied in public or non-public spaces is determined by people’s needs for privacy and security. Facial recognition in public or non-public spaces are more likely to be accepted when there are outstanding security concerns, whereas its acceptability is lower when people would give more priority to their privacy as security concerns are not prominent.
Conclusions and Suggestions

Owing to the research method adopted by the team, the proportion of respondents with a bachelor’s degree or higher is 83.6%. In this sense, the survey results may be more of a reflection of the opinions of individuals with higher educational background. Also, certain additional conditions are applied to a few questions, which may affect the universality of the conclusions to some degree.

This survey was conducted during the COVID-19 pandemic. China has taken proactive measures to check body temperatures and track close contacts with facial recognition in response to this public health crisis. Although it is a densely populated country, China only spent around 2 months to bring the pandemic within the “mitigation stage.” In this process, the Chinese people also learned about the positive role of facial recognition and other related technologies. We can see that the respondents, while acknowledging the positive role of facial recognition in responding to a public health crisis, also express concerns about issues such as privacy and security. They hope that the usage of facial recognition can better protect personal privacy and data, safeguard public security, and also better leverage technological potential in the fight against public health crises in compliance with laws and regulations. Some respondents remain neutral on or disapproving of reducing unnecessary applications of facial recognition, which reveals their concerns about the recurrence of crises or emergence of new crises even when the pandemic is over, and their expectations for the potential uses of these technologies in preventative measures. The pandemic itself and this survey both show that strengthening a technology-driven early warning and prevention system for public health crises is all the more necessary, so as to prevent the outbreak of another large-scale public health crisis. Additionally, some respondents of this survey also express their hope to enhance public education about facial recognition and thus improve the public understanding of the technology itself, as well as its potential applications. This will encourage more people to accept and support the “proper use” of facial recognition and other AI technologies, participate in the supervision of potential technological or ethical risks arising from technological development and application, and offer feedback and suggestions.