DL Seminar | Alondra Nelson - I Contain Multitudes

By Willy Lin | MS Student, Cornell Tech

Voluntarily supplied consumer data is entering modern industry in droves. With it comes extreme data privacy issues and unseen aftereffects. Companies focused on the efficacy of data analysis and manipulation commonly fail to properly characterize the effect of such user-driven data. In her DLI Seminar talk on September 12, 2019, Professor Alondra Nelson, Faculty at Princeton's Institute for Advanced Study and President of the Social Science Research Council, uses the case study of direct to consumer (“DTC”) genetics as a way of showcasing the shortcomings of current data ethics and policy.

Genetic Data

Nelson began her talk explaining how genetic data could be used as a standpoint from which to question issues around data privacy and ethics by examining its effects on social inequality and dispossessed communities. The essential notion within data privacy is that data that represents individuals always implicates others. Within bioethics, medical information taken about an individual can inform and implicate other people within similar sets. The use of genetic data, therefore, has to be coupled with the consideration of social implications well beyond intent. Data without ethical barriers can be reidentified, so in the interest of social good, professional norms can and should be created to prevent this. This notion applies to data biasing as well, such as the creation of social queries that take into place these ethics instead of just aggregating binary data.

Nelson classified different boundaries within genetic data and testing into three distinct categories:

1. Domains and institutions;

2. Regulatory schema like bioethics;

3. Micro in practice.

Within domains and institutions, in 2008, the American Society of Human Genetics noted that guidelines should be developed for counseling ancestry estimation within the health care setting. Many criminal cases throughout the years have utilized genetic data as a way to arrest suspects. Nelson noted that in the BTK Killer case police utilized their own databases and protocols to do familial searching. Police examined the DNA on a letter that Dennis Rader sent. They cross examined this with a local medical center and identified through a Pap smear a familial relationship. The killer was implicated through this and other corroborating evidence.

However, for every positive result of DNA testing and genetic databasing, there exist other false positives with real negative consequences. The case of Michael Usry, Jr. details one such event. Detectives investigating the 1996 cold case murder of Angie Dodge identified a familial match with Usry, Jr.’s father who had submitted his DNA information through a nonprofit which was later purchased by Ancestry.com. Police investigated and interrogated Usry, Jr. for 6 months before further DNA testing proved that he was the wrong person. Although this case was publicized, there are no articles detailing the potentially thousands more individuals investigated and who had their DNA analyzed by law enforcement. There are no laws and regulations in this data management.

Regulation of DTC Genetic Testing

While regulations for the laboratories analyzing DTC genetic data, no state-level ethical regulations exist. For instance, 23andMe, a personal genomics and biotech company, is currently not classified as a medical entity that can be subject to the strict government regulations within the medical space. The state of California allows these companies to proceed without restrictive regulations for a portion of their tests. Many similar companies that conduct non-medical testing fall through the cracks of these agencies. Other regulations that could apply include HIPAA, which guarantees that patients’ genetic test results will be private and protects them from potential discrimination. However, DTC testing is not considered a covered entity. The Genetic Information Nondiscrimination Act of 2008 prohibits types of genetic discrimination related to medical genetics. However, insurance is not considered medical and is not covered.

In practice, DTC genetic testing companies utilize unique tactics to expand their genetic database and incentivize consumers to willingly provide their genetic information. Nelson describes the case of Sara, who provided her DNA to a DTC genetic company because she wanted to understand her ancestry and identify health concerns within her family tree. Another person, Martin, became engrossed in genetic study after submitting his information too. This halo effect of genetics is what DTC companies use to market their services and, as a result, build out gaps within their genetic databases. These gaps tend to be associated with people of color and minorities. An Ancestry.com marketing tactic involved promoting the possibility that a customer could belong to an African diaspora group, incentivizing African Americans to provide their DNA to Ancestry.com, thus making it available to use by governmental agencies.

Conclusions

Ultimately, genetic testing and analysis provides great utility to both companies and individuals who want to learn more about their history and health. There are many hidden caveats that apply to submitting information to a DTC genetic company. Often, these companies profit through the packaging and wholesale sale of this genetic information to industries such as pharmaceutical and medical insurance where the information can be used to negatively affect a multitude of individuals. By providing this data, individuals can also implicate other people in similar genetic circles. The Innocence Project is a non-profit organization that has exonerated over 300 people that were falsely accused and convicted with genetic evidence. This could not have been achieved without the efforts of amazing men and women who understand the potential pitfalls of unregulated genetic databases, and was necessitated as a result of being in a society that incarcerates those people in the first place.

Genetic data contains faceted information. It is multivocal and can be used in genealogical and medical and forensic cases. Genes are an omnibus of a person’s biological identity and contain many types of information simultaneously. With such a great and multitudinous treasure trove of information, data ethics and policies should reflect the magnitude of implications that data brings with it.