is now a condition of most public research funding.
The argument is strong: publicly funded evidence should be available for checking and reuse.
depends on it, and so does the detection of error that peer review cannot reach.
The people who supplied the data were promised something different.
A participant to a study with a stated purpose, a named team and a defined end date.
Open converts that into an indefinite permission for unknown purposes.
was the answer for twenty years and is no longer sufficient.
Repeated demonstrations have shown that a small number of variables identifies most individuals in a population.
A date of birth, a postcode and a diagnosis are frequently enough.
Genetic data cannot be anonymised at all, since the is the data.
The field has moved from open publication to controlled access in response.
A researcher applies, states a purpose and works inside an environment that the data never leaves.
That preserves most of the scientific benefit and creates a new with its own incentives.
Access committees have refused applications for reasons that look institutional rather than ethical.
Publishing the refusals, with reasons, is the cheapest correction available and is rarely done.
Participants themselves are almost never part of this .
The few studies that placed participants on the access committee report that they approve more applications than researchers do, not fewer.
The assumption that the public is the obstacle to data sharing has survived a long time without evidence.