There are calls to stop Australian police from using a new technology it claims can predict what someone looks like — down to ancestry and facial features — for fears that it will be used for racial profiling.
On Sunday the Australian Federal Police announced it has begun using “Massively Parallel Sequencing” (MPS) as part of its criminal investigations.
“The new technology … can provide predictions for visual traits of criminals from the DNA they leave at a crime scene allowing investigators to predict gender, biogeographical ancestry, eye colour and, in coming months, hair colour,” it said.
Two Australian academics, the Australian National University’s Dr Jenny Davis and Monash University’s Jathan Sadowski, want the use of this new technology halted pending further consultation with independent experts and communities.
In an open letter to the AFP, Davis and Sadowski outlined concerns that the technology would be used as a justification for racial profiling, while also noting that crime prediction tools have racial biases built into them. The pair has published the letter online, allowing others to add their names in support.
“We are concerned about the negative effects MPS will have on marginalised communities and on the credibility of the policing system,” they wrote.
Davis tells Crikey she was astounded to see the AFP say it was using the nascent technology. She says policing has a history of embracing “objective” technologies that end up reinforcing racism because of inbuilt assumptions.
“There’s always been a tie between race, racism and policing,” she said.
A 2020 paper published in the Australian Journal of Forensic Sciences about MPS documented a trial of Queensland detectives using the technology on four real criminal cases. In the workshops, detectives excluded 30% of suspects based on the technology but noted that participants frequently made errors.
“There was a high rate of correct interpretation and use of the MPS results; however, the frequency of errors suggest that detectives need support from a trained forensic biologist to ensure accurate result interpretation,” the paper’s authors wrote.
AFP has been contacted for comment about how it assessed MPS and how it intends to safeguard genetic information.
Wouldn’t this have the opposite effect? If police are currently pressuring certain groups due to unconscious bias and MPS eliminates those groups as suspects then wouldn’t police be less likely to go on fishing expeditions in those communities?
Correct
Isn’t it similar to the current ‘Caucasian’, or ‘of African appearance’, but with the potential to be more accurate?
Cam, when I was a kid back in the 1950’s and 60’s we used to have the Liberals and political wing of the Catholic Church, aka the DLP, running around looking for communists under every bed. (How I ever escaped that dragnet I will never know!!)
Nowadays, it is the crazies on the left who leave no stone unturned in their fanatical quest to find ‘racists’ and ‘xenophobes’, etc. hiding under every cabbage leaf.
I had a look at the letter that these two intrepid academics have sent to the AFP and I am entirely unconvinced by their arguments. I notice too that neither of them is likely to have any background in genetics or criminology.
A quick check on the internet reveals that Dr Jenny Davis is a sociologist who “focuses on role-taking, status, stigma, and identity, along with technological affordances and the politics of digital design”. Whatever that all entails is entirely unclear to me but I feel that criminology and genetics are not a big part of it.
Projects of interest to Jathan Sadowski, on the other hand include studying how people live in smart homes, reshaping how utility companies forecast digital energy futures, and analysing the implications of automated decision-making in society. As an interdisciplinary scholar, Jathan’s research sits at the intersection of technology studies, political economy, and urban geography.
How the professional roles of these two ties in with apprehending criminals is a complete mystery to me.
I am fully in favor of any reasonable approach to apprehending criminals. I am not convinced that Massively Parallel Sequencing technology represents any problem at all. If it can be used to identify the racial or national background of any ‘person of interest’ in a criminal investigation then I am fully in favor of it. As fate has it, I have a Caucasian background. If the AFP were to identify any suspect as being of “Caucasian background”, am I really supposed to feel victimized or afraid? What nonsense!!
As with any new approach there may be mistakes made but I am sure that these are also made with fingerprint and other crime-fighting technologies as well.
Nice researching Bert.
I dont get it. There is a technology that can OBJECTIVELY identify the (probable) visual characteristics of a perpetrator. It doesn’t violate any privacy rules. All it does is give police more leads and narrow down the range of suspects. How on earth is that in of itself a bad thing. This could help catch more rapists or child abductors.
If police use the results of the test to harass a subset of society, that’s got nothing to do with the technology. This is like saying CCTV should have a filter that prevents police seeing the skin color of a perpetrator.
Lets be honest, if the test was only able to identify people who where white, male and conservative, Crikey would be running a 10 part series on how this is the most important development in policy this century. The left would be demanding it is adopted to protect the vulnerable. Feminists would be demanding it to protect women. Rah rah rah.
*policing*
How accurate is it – that it won’t result in wrongful pursuits? ….. Not another tool for AFP to misuse? What would it have done to the Bali 9? ….. Helped in the arrest of leakers against the Coalition government?
…. I’ve often wondered what would happen, if a perpetrator bagged the clippings from a barber’s shop floor and scattered the contents around the scene of their subsequent crime.