I endorse Emma Kowal and colleagues' call for more-responsible sharing of people's research data (Nature 546, 474; 2017). The reuse of data without participants' consent is a serious ethical problem and could discourage participation in future projects (see, for example, go.nature.com/2ub6qrq).

To the chagrin of privacy advocates, the US Department of Health and Human Services introduced reforms in January that allow sharing of medical-research data without re-consent by patients (see Nature 541, 449; 2017). Researchers must now share their archived data, which were collected before the National Institutes of Health (NIH) data-sharing policy of 2003 and when consent forms did not let people opt out of having their data shared. Researchers who fail to comply can have their NIH funding withheld (see go.nature.com/2vzjthi).

This policy is particularly problematic for anonymous donors who have stigmatizing conditions such as alcohol or drug addiction. De-identified data in family collections can now be identified using genealogy databases (J. Bohannon Science 339, 262; 2013). With the hacking of medical records for profit becoming more pervasive in the United States (see go.nature.com/2uytcpg), people who donated their private information could now be compromised.