"You have nothing to fear if you're not going to be a criminal," said a spokesperson for the Bensalem, Pennsylvania police.
Bensalem is one of a growing number of police departments that have adopted Rapid DNA technology, a DNA sequencing machine that officers refer to as "the magic box." In the past, DNA testing was only performed by accredited labs, but now police can do it without assistance -- or accreditation.
The New York Times interviewed a detective from Bensalem who began using a Rapid DNA testing machine after a few hours of manufacturer training.
"There really are no actual rules written anywhere," he told the Times.
DNA samples that police collect are currently uploaded only into county databases, which are much more loosely regulated than the FBI's nationwide Codis database. However, the FBI is currently working on allowing some police departments to upload some genetic profiles to Codis.
The profiles would have to be extracted from cheek swabs with a single person's DNA. Most scientists agree that Rapid DNA testing devices are most reliable when used in this way. They were not designed to handle crime scene evidence, numerous scientists told the Times.
Yet increasingly, police are using the machines for crime scene evidence, such as cigarette butts and discarded pop cans.
They defend this practice in part by saying that the technique is only used to generate leads -- the evidence is almost never introduced in court. But that could still mean that unreliable evidence is being used to prosecute people. It could also reinforce stereotypes, as DNA is being collected from people deemed suspicious by the police.
Last January, the National District Attorneys Association issued a statement saying it "does not support the use of Rapid DNA technology for crime-scene DNA samples unless the samples are analyzed by experienced DNA analysts."
Science fiction meets the wild, wild west
Problems have already cropped up elsewhere in the world. In 2017, the Swedish Forensic Center began and then prematurely ended its trial of Rapid DNA. One reason was that almost 25 percent of the samples it tested resulted in no usable profile -- and the sample was consumed by the test.
Additionally, a great many samples resulted in a faulty profile with no error messages or warnings. If the center hadn't been manually reviewing the results, these faulty profiles might have been used in real cases.
A New York University law professor is critical of the whole idea. An approach that "starts with everybody's a suspect, and then let's go see if we can find a crime they've committed -- I think that's a deeply problematic inversion of how we do things," she said.