Contact tracing has been a key measure in an attempt to slow the spread of COVID-19. Many countries, including the United States, have used contact tracing to track and control the spread of the disease. The measure has been touted as a success in the states and countries that managed to implement a well-functioning system with adequate preparation, testing and tracking, welfare support, and effective leadership. However not all contact tracing efforts were successful. The United Kingdom’s contact tracing efforts have been noted to be a failure, with the country’s government scientific advisory group reporting that the program had a “marginal impact on transmission.” The United States has similarly experienced failures in several states. Some of the reasons that programs fail include lack of local support, sheer number of cases, and delays in testing and identification. Even strong advocates for robust contact tracing programs have conceded that “it is impossible to do meaningful or substantial contact tracing with huge numbers of cases.” Shortage in personnel performing contact tracing also likely contributes to the difficulty.
One key tool that could have mitigated some of the problems with contact tracing is the use of digital contact tracing. In fact, several countries, notably South Korea and Singapore, have successfully implemented digital contact tracing programs. Apple and Google partnered to create a COVID-19 contact tracing technology which utilizes Bluetooth technology. Claims have been made that implementing AI mechanisms in such technology would further increase the effectiveness of Bluetooth contact tracing by boosting the ability of the phones to detect nearby phones, which remains shaky with just the use of Bluetooth technology. Pairing Bluetooth contact tracing with AI learning mechanisms could help establish an effective contact tracing system that would help mitigate the spread of a disease in future pandemics. Additionally, a startup is developing a system that would use AI to analyze non-facial attributes of a person in a video for contact tracing purposes. However, in order for such contact-tracing mechanism to work, the government would need to convince the public to actually use the technology. This could prove to be a difficult task – digital contact tracing saw little success in the United States due to strong privacy concerns. Several states have also launched applications relying on GPS, but these apps have not been widely adopted. There have also been legal challenges to contact tracing programs, which argued several violations of the Bill of Rights, notably the Fourth Amendment. Combine this with privacy issues surrounding the use of AI and Big Data, and AI-integrated contact tracing would likely encounter major privacy hurdles before it could be successfully implemented in the United States.
The digital contact tracing efforts currently adopted by the U.S. carefully balance the protection of privacy with effectiveness. The current limited scope of privacy laws means that “there is no baseline data protection law that would protect the sensitive data” obtained through the contact tracing apps. The protection offered for consumer data comes from a “patchwork of” several different federal laws. Examples of relevant laws include the Health Insurance Portability and Accountability Act’s data protection requirements, the Communications Act, Family Educational Rights and Privacy Act, the Electronic Communications Privacy Act, etc. The fragmentation of data protection laws means that without laws ensuring uniformity in safeguarding individual data, not all contact-tracing data will be protected. Lack of comprehensive protection measures, combined with American skepticism towards data collection done both by big tech companies and by the government, make it extremely difficult to get public buy-in to contact tracing mechanisms. Additionally, even when a more comprehensive legal framework for data protection measures can be established, the U.S. government and the Courts would also need to address other individual rights claims such as the 4th Amendment claim. Adding Artificial Intelligence to this framework would further exacerbate this problem, as it would add another wrinkle to the consideration of the applicable laws. Implementation of AI in such a system would increase the available data that could be under the protection of such laws. Additionally, public concern for privacy in Artificial Intelligence has often been highlighted and could serve to deepen the resistance that individuals already feel to using these measures. Especially considering the problem of bias and racism in AI technology, such concerns are not unfounded.
Integration of AI and digital contact tracing would require a major change in the current data privacy law and public attitude towards privacy. Whether benefits from digital and AI contact tracing warrant such uprooting of current data protection framework remains to be debated.
* Saika Suzuki is an Associate Editor on the Michigan Technology Law Review.