Among direct-to-consumer relative-finder participants who learned new information about themselves or their relatives, approximately 11% reported decisional regret, according to Guerrini et al. (2022) in the American Journal of Human Genetics (n=23,196; 61% learned something new, 11% of those scored above the Decision Regret Scale cutoff, and 2% reported net-negative consequences). Learning something new can mean discovery shock from unexpected paternity revelations (a “Not Parent Expected” result), identification of unknown biological relatives, or unanticipated high-penetrance disease risk variants such as BRCA1/2. This 11% applies to the subset who receive genuinely unexpected findings, not to all test takers — the majority of people who test report positive or neutral experiences (ancestry discovery, confirmation of family stories, actionable health information), and the donor-conceived subgroup reported the highest regret of all. The inaction side is harder to measure directly because no survey asks non-testers whether they regret opting out. The closest proxy comes from an AP-NORC Center for Public Affairs Research poll (June 2018, n=1,109): 17% of US adults had been genetically tested and another 52% said they were interested in having it done, primarily citing curiosity about ancestry and the desire to understand genetic health risks. Interest is a latent signal, not regret — many of those who are “interested” may never act, and interest can coexist with no felt regret about not having tested yet.
The decision’s structure is unusual because its consequences are irreversible in only one direction. A person who tests and discovers distressing information cannot un-know it; the cascade of downstream decisions triggered by unexpected results — preventive surgeries, family disclosures, identity reckonings — carries its own regret profile independent of the original test decision. A person who opts out retains the option to test later, meaning inaction here is not a permanent foregone opportunity but a deferral. This structural asymmetry likely suppresses inaction regret and amplifies action regret relative to other regret pairs.
The risk calculus shifted substantially after October 2023, when a credential-stuffing attack on 23andMe exposed the genetic ancestry data of approximately 6.9 million users, followed by 23andMe’s bankruptcy filing in March 2025 and the subsequent uncertainty about who would control and monetize stored genetic profiles. Users who tested between 2017 and 2023 made their decision under a different risk profile than the one that materialized. The pre-breach AP-NORC finding that 50% of US adults were already extremely or very concerned that for-profit DNA companies would share their genetic information proved prescient: a large share of non-testers appear to have anticipated a data-handling risk that materialized dramatically for millions who had already tested.