Academic and regulatory discussions surrounding the data quality principle have almost entirely focused on the threats of low accuracy of personal data, having lost sight of the potential risks arising from the undesirable high level of data accuracy. Today’s increasing level of correctness and precision in data processing will bring forth a series of new challenges, including new forms of discrimination and the loss of individual manoeuvre space, which in turn call for the preservation of a socially acceptable range of data accuracy. While the current EU data protection framework, characterised by the principles of data minimisation and purpose limitation, may provide certain safeguards against such new threats, some flaws remain. A robust solution would therefore entail an updated understanding of the data quality principle, which covers both extremes of the data accuracy spectrum, and acknowledges that data quality is an ethical issue to be decided by both the society and individuals.