Simoncelli argues that first, preserving non-felons' DNA "turns the presumption of innocence on its head," turning anyone in the database into a suspect. Even convicted felons aren't automatically guilty of future crimes. This is anti-democratic in nature, and dangerous in practice.
Second, at least in the American system, institutional safeguards enshrined in the Fourth Amendment would be threatened by DNA databases.
Regardless of whether a DNA bank should be considered beyond the general needs of law enforcement, the proposition that the government's "special needs" outweigh the privacy interests of innocent persons seems beyond the pale, as a matter of Constitutional principle. While it is plausible that the courts could uphold the forcible taking and analysis of DNA of persons arrested on the basis of some diminished expectation of privacy while in confinement, the permanent retention of that DNA cannot be justified on this basis unless a suspect is convicted of a crime.Beyond 4th Amendment considerations, DNA databases create unique privacy concerns.
Unlike fingerprints - two-dimensional representations of the physical attributes of our fingertips that can only be used for identification - DNA samples can provide insights into personal family relationships, disease predisposition, physical attributes, and ancestry. Such information could be used in sinister ways and may include things the person herself does not wish to know.Abuse of such a system is highly likely.
[S]pecific cases of abuse of police databases indicate that penalties alone do not sufficiently deter misuse. In 2001, it was revealed that more than ninety known cases of abuse of Michigan's Law Enforcement Information Network had occurred over five years. Abuses included police officers and other law enforcement personnel tapping into the network to obtain home addresses or other background information on love interests and seeking revenge or an upper hand in personal, legal or political conflicts. And while Michigan law clearly indicates that such an abuse qualifies as a misdemeanor, punishable by up to ninety days in jail and a $500 fine upon conviction, only three of the officers were prosecuted for these crimes.Simoncelli details many practical concerns that are of secondary concern here, given that their impact is utilitarian rather than a matter of violated rights. They include the diminishing returns of an expanded dataset, the possibility of false convictions via planted DNA evidence (the paradoxical result of heightened trust in such evidence), overworked crime labs, untold costs (somewhat mitigated by falling prices), and the necessity of a total-population database to ward off concerns about "racial distortions in our criminal justice system."
In all, the article is well worth reading as a primer on some of the primary arguments in the debate.