Voluntariness at the time one provides data is an important, overlooked part of providing informed consent. In medicine, informed consent requires voluntariness, yet the on-the-ground experience may reveal pressures to comply. The new landscape of responsible technology, while it incorporates certain types of consent like clicking to accept cookies, needs more definition and clarity around voluntariness. A new threshold for voluntariness would improve the ability to keep things private by distinguishing quid pro quos from times that people voluntarily wish to provide information. In medicine, voluntariness is part of informed consent, yet it is often not meaningfully achieved.
At its best, voluntariness means there are no strings attached, that information is freely provided. Data is itself not a neutral thing. For example, to make a doctor’s appointment, I must give some insurance information. Many doctors will not allow you to schedule and have no ability to price their services or products until they know which insurer will be paying. Is it voluntary if it is in exchange for the privilege of making the appointment?
Some people do not want to divulge their weight, even to the doctor. While it is medically relevant, hopping on the scale, creating a data-bite, is not exactly voluntary. It seems required by a doctor or nurse who appears to be in a position of authority. Many people do not understand their right to opt out of providing various types of information in the medical context, an important form of refusal.
Principles of Choice and Consent in Privacy
Some privacy frameworks refer to “choice” as a principle, often also “choice and consent”. Choice supports systems that include decisions about whether to accept cookies, but choice does not speak explicitly to voluntariness. Some choices are futile – if people are barred from access to websites or from choosing certain medical treatments, then they may feel there is no choice regardless of providing informed consent that is defined as voluntary. Many privacy frameworks behind technology ethics, international organizations’ privacy guidelines, medicine, and consumer data laws refer to either choice or consent, but the scales with which to measure voluntariness are missing. The quid pro quos range: in exchange for chemotherapy, people sign off and consent to take on huge risks; in accepting cookies people allow companies to track some of their browsing habits and data in exchange for use of websites. While different, both have an element of pressure, undermining voluntariness.
Adding a Voluntariness Principle
Transparency, fairness, respect, use limitation, collection limitation, choice, and consent, and the other principles often associated with technology do not cover voluntariness adequately. A voluntariness principle would shift the attention to the person providing personal data and the point of collection. The lived experience of those filling out “paperwork” in a doctor’s office or clicking to allow cookies, anonymized data, or deidentified medical or health data reflects that consent is a box to check, often seen as a way that doctors avoid liability.
While clearly people consent to medical treatments and acknowledge risk commonly, there are also many times they experience duress and feel like they have no real choice. Consent is the cost of participation. Rather than assuming people voluntarily consented to providing data, the burden could shift. It seems possible that informed is more prominent than voluntary in the characteristics of consent surrounding data, tech, or medical treatment. Consent forms center on an understanding of treatments and risks over an understanding of rights to refuse or voluntariness. The ability of doctors to exert power or the patient perception of physician power, and the ability of websites to create required fields that one must fill in to proceed undermine true voluntariness. A scale of voluntariness and an exhaustive list of considerations would help consumers of health care and other goods and services reflect on the differences between voluntariness and forced acceptance of conditions of participation that the consumers did not take part in shaping.
Example: Facial Recognition Technology and Discrimination
A doctor uses facial recognition technology to determine a risk for a genetic debilitating disease. The person seeking care for a headache signed many documents, some HIPAA, some consents to have an MRI that is also corroborated with the facial recognition technology. A boilerplate consent form said some MRIs are combined with facial recognition technology software. The paperwork said deidentified MRIs may be used for research purposes. The patient supplied insurance data prior to the appointment. The insurer finds out the genetic diagnosis because the hospital in good faith seeking reimbursement copies the insurer on the facial recognition technology report. The insurer drops the patient or charges more claiming there is a preexisting condition. GINA does not cover genetic information gathered this way so it would not prevent genetic discrimination in this case.
Some thoughts: Whether consent was voluntary depends on both an understanding of the complexity and the breadth and ability of FRT to make a genetic diagnosis, the on-the-ground pressures of just signing documents when you need an appointment or MRI, and the inability to access diagnostic tools or treatments without signing papers, giving insurance information, or creating a profile. Are there considerations, a measurement or scale, or a best definition of voluntariness?