Parabon Nanolabs/Snapshot Accused of Racism

[HmmMysterious]

This is a snapshot created by Parabon Nanolabs on October 4, 2022 for the Edmonton Police Service. The snapshot caused a huge amount of outrage when the Edmonton Police posted it on their Twitter account. The police were later forced to shut off comments.

https://twitter.com/x/status/1577296746698661890

Dr. Adam Rutherford on Twitter said, "Geneticist here. You can't make facial profiles or accurate pigmentation predictions from DNA, and this is dangerous snake oil."

I agree with critics and Dr. Rutherford. This Snapshot reeks of racial profiling. Corina Newsome, M.Sc. sums up my thoughts perfectly when she said, "And so now every black man who looks even marginally like this Al-generated image will be a target. And y'all couldn't care less. We cannot afford to have this racism and incompetence directing our science/tech "advancements".

Honestly the other Snapshots Parabon Nanolabs have made aren't even that good. The program should fold. What are your thoughts on this?

 

[Betty P]

View attachment 371139

This is a snapshot created by Parabon Nanolabs on October 4, 2022 for the Edmonton Police Service. The snapshot caused a huge amount of outrage when the Edmonton Police posted it on their Twitter account. The police were later forced to shut off comments.

https://twitter.com/x/status/1577296746698661890

Dr. Adam Rutherford on Twitter said, "Geneticist here. You can't make facial profiles or accurate pigmentation predictions from DNA, and this is dangerous snake oil."

I agree with critics and Dr. Rutherford. This Snapshot reeks of racial profiling. Corina Newsome, M.Sc. sums up my thoughts perfectly when she said, "And so now every black man who looks even marginally like this Al-generated image will be a target. And y'all couldn't care less. We cannot afford to have this racism and incompetence directing our science/tech "advancements".

Honestly the other Snapshots Parabon Nanolabs have made aren't even that good. The program should fold. What are your thoughts on this?

Does the same accusation apply for Caucasian men whose profiles they developed based on the perp's genetic phenotype have also been discriminated against? If you can't make accurate predictions of facial profiles or pigmentation from DNA, has Parabon also been wrong about the physical characteristics they developed for Caucasian DNA?

I have to admit, some of the profiles they've done for Caucasian persons have a sort of generic look to them, but others seem fairly accurate.

Here's a paper published in 2019 about guidelines to follow when using DNA phenotyping.
Studies have shown the three parts of DNA phenotyping do have some positive predictive value, but must be used carefully.

The Use of Forensic DNA Phenotyping in Predicting Appearance and Biogeographic Ancestry

Persons whose identifying DNA profile (STR profile) is not yet known to the ingvestigating authorities cannot be identified by standard forensic DNA analysis (STR profiling) as it is now practiced. In view of the current public debate, particularly in ...

www.ncbi.nlm.nih.gov

 

[HmmMysterious]

Does the same accusation apply for Caucasian men whose profiles they developed based on the perp's genetic phenotype have also been discriminated against? If you can't make accurate predictions of facial profiles or pigmentation from DNA, has Parabon also been wrong about the physical characteristics they developed for Caucasian DNA?

I have to admit, some of the profiles they've done for Caucasian persons have a sort of generic look to them, but others seem fairly accurate.

Here's a paper published in 2019 about guidelines to follow when using DNA phenotyping

The Use of Forensic DNA Phenotyping in Predicting Appearance and Biogeographic Ancestry

Persons whose identifying DNA profile (STR profile) is not yet known to the ingvestigating authorities cannot be identified by standard forensic DNA analysis (STR profiling) as it is now practiced. In view of the current public debate, particularly in ...

www.ncbi.nlm.nih.gov

Yes Parabon is wrong because they just decide to portray suspects with average BMI and at 25 years old. The program honestly does more harm than good. One of their Snapshots featuring a Caucasian suspect was dragged because of these reasons.

https://twitter.com/x/status/1577689358500511746

 

[Betty P]

Yes Parabon is wrong because they just decide to portray suspects with average BMI and at 25 years old. The program honestly does more harm than good. One of their Snapshots featuring a Caucasian suspect was dragged because of these reasons.

https://twitter.com/x/status/1577689358500511746

They don't have much choice but to use averages for weight. I've seen them issue Snapshots of a suspect's possible appearance at various ages, with different hair styles, with and without facial hair, etc. They go to great detail to explain their Snapshots are simply an estimation of what the perp looked like.

If a Snapshot of a Caucasian suspect was criticized, then it seems hard to call it a racial issue. When Snapshot suspects have been identified in the past, they usually look very similar to Parabon's estimate. Not always, but often. It's a tool that many law enforcement agencies use as a last resort, to generate tips from the public when they have little else to work with.

 

[evilwise]

Here is a link to all of Parabon's phenotype facial reconstructions, both unsolved and solved with photos of the person attached.

Police Predictions - Parabon® Snapshot® DNA Analysis Service

snapshot.parabon-nanolabs.com

In my opinion there is little room for debate that this is an invaluable tool but it's true that science is racist. We'll see how that reality shakes out over the next decade or so.

 

[Betty P]

Here is a link to all of Parabon's phenotype facial reconstructions, both unsolved and solved with photos of the person attached.

Police Predictions - Parabon® Snapshot® DNA Analysis Service

snapshot.parabon-nanolabs.com

In my opinion there is little room for debate that this is an invaluable tool but it's true that science is racist. We'll see how that reality shakes out over the next decade or so.

JMO, the hairstyles and BMI are always the hardest to predict.

 

[evilwise]

JMO, the hairstyles and BMI are always the hardest to predict.

Age too, which affects a person's hair. I'm pretty sure that Parabon presents a number of various composites for any given case that somewhat cover different looks that people with the same bone structure and pigmentation might have.

 

[MadMcGoo]

Thanks for the link @evilwise . I agree the tool is invaluable! It beats relying on a decades-old clay bust reconstruction IMO. But they do usually give the disclaimer about weight, hair style, etc.

I found this when looking into the OP and I very humbly admit that I didn’t know Parabon was a catalyst to the changes in GEDmatch.

The controversial company using DNA to sketch the faces of criminals (BBM)

The company had made its name by comparing suspects’ DNA to profiles on genealogy databases and piecing together family trees to track down alleged offenders.

But then controversy erupted over a case Parabon helped to solve, in which a teenage boy had violently assaulted a septuagenarian in a Mormon meeting house in Utah. The Utah case generated public backlash because of concerns over privacy.

Genealogists at Parabon had been generating leads by sifting through a database of DNA tests called GEDMatch, a free-to-use website that allows users to upload test results in the hope of finding long-lost relatives. At the time, GEDMatch allowed law-enforcement agencies access to the profiles to help solve murders and sexual assaults, unless users specifically opted out. The police, aided by Parabon and companies like it, made new arrests weekly.

But the Utah case was not a murder or a sexual assault — and so was not covered by the website’s disclaimer. The assailant had left traces of blood at the scene, and the detective in charge of the case, Mark Taggart, made a personal plea to GEDMatch’s founder, Curtis Rogers, for access to the database. When it was granted, Parabon, which had initially refused the case, signed on. The company traced several partial DNA matches to individuals living in the area, and narrowed in on a suspect, a teenaged boy who was a relative of one of them. Taggart made an arrest.

That triggered an immediate backlash from genealogists, privacy experts and the wider public at the violation of GEDMatch’s agreement with its users. In response, Rogers required the site’s millions of users to specifically opt in to law-enforcement use. Overnight, Parabon lost its major source of DNA data.

 

[HmmMysterious]

Here is a link to all of Parabon's phenotype facial reconstructions, both unsolved and solved with photos of the person attached.

Police Predictions - Parabon® Snapshot® DNA Analysis Service

snapshot.parabon-nanolabs.com

In my opinion there is little room for debate that this is an invaluable tool but it's true that science is racist. We'll see how that reality shakes out over the next decade or so.

I don't think so and I think more geneticists besides Dr. Adam Rutherford will also trash it.

 

[bombardier]

Yes Parabon is wrong because they just decide to portray suspects with average BMI and at 25 years old. The program honestly does more harm than good. One of their Snapshots featuring a Caucasian suspect was dragged because of these reasons.

https://twitter.com/x/status/1577689358500511746

That comparison on the bottom (of convicted murderer Robert Dwane Washburn) is not really fair; they were trying to portray the perp as he would have looked at the time of the crime, which was 30 years prior. If you look up his old mugshots, he did look similar to the composite at the time.

Honestly I don't agree with police using Snapshots to look for public tips about a suspect, regardless of race. The composites are, of course, generic, and many many people could potentially fit. I have to imagine they have to deal with an influx of tips based entirely on look-alikes, but if they weren't getting any tips in the first place, I can understand why this might be better. The Snapshots are going to be even MORE generic when constructing composites of black or asian people due to less variation in hair and eye color, so I understand the complaints that some people are making.
All in all though I think it's a very interesting program with some really great success stories, and I have no problem with its use in other investigations/cases.

 

[Betty P]

Thanks for the link @evilwise . I agree the tool is invaluable! It beats relying on a decades-old clay bust reconstruction IMO. But they do usually give the disclaimer about weight, hair style, etc.

I found this when looking into the OP and I very humbly admit that I didn’t know Parabon was a catalyst to the changes in GEDmatch.

The controversial company using DNA to sketch the faces of criminals (BBM)

The company had made its name by comparing suspects’ DNA to profiles on genealogy databases and piecing together family trees to track down alleged offenders.

But then controversy erupted over a case Parabon helped to solve, in which a teenage boy had violently assaulted a septuagenarian in a Mormon meeting house in Utah. The Utah case generated public backlash because of concerns over privacy.

Genealogists at Parabon had been generating leads by sifting through a database of DNA tests called GEDMatch, a free-to-use website that allows users to upload test results in the hope of finding long-lost relatives. At the time, GEDMatch allowed law-enforcement agencies access to the profiles to help solve murders and sexual assaults, unless users specifically opted out. The police, aided by Parabon and companies like it, made new arrests weekly.

But the Utah case was not a murder or a sexual assault — and so was not covered by the website’s disclaimer. The assailant had left traces of blood at the scene, and the detective in charge of the case, Mark Taggart, made a personal plea to GEDMatch’s founder, Curtis Rogers, for access to the database. When it was granted, Parabon, which had initially refused the case, signed on. The company traced several partial DNA matches to individuals living in the area, and narrowed in on a suspect, a teenaged boy who was a relative of one of them. Taggart made an arrest.

That triggered an immediate backlash from genealogists, privacy experts and the wider public at the violation of GEDMatch’s agreement with its users. In response, Rogers required the site’s millions of users to specifically opt in to law-enforcement use. Overnight, Parabon lost its major source of DNA data.

Agree, LE probably shouldn't make arrests based on partial DNA matches, unless they have some other substantial corroborating evidence.

Even when making arrests based on a full DNA match, LE has to make sure the evidence they're using for DNA actually came from the perpetrator of the crime. Some law enforcement agencies apparently have to learn from their mistakes.
The US isn't the only country struggling with using DNA evidence ethically and responsibly. A lot of European countries have had similar debates.

 

[MadMcGoo]

It’s like freckles, which I see often on Parabons snapshots. They can occur in anyone but most commonly don’t. Their images/recons seem largely based on science and math. JMO

 

[bombardier]

Here is a link to all of Parabon's phenotype facial reconstructions, both unsolved and solved with photos of the person attached.

Police Predictions - Parabon® Snapshot® DNA Analysis Service

snapshot.parabon-nanolabs.com

In my opinion there is little room for debate that this is an invaluable tool but it's true that science is racist. We'll see how that reality shakes out over the next decade or so.

Just to back up my opinion, I looked at the posters pertaining to [ARRESTED] or [GUILTY] perps. I know that Parabon is just using those as an example of their successful AI technology, but imo it's a bit misleading because it almost implies that the creation of the snapshot composite LED to the arrest of that person (or I thought that's what it implied, lol). Clicking on the links and reading the articles, majority of those cases were solved through genetic genealogy (often thru Parabon as well) or other means. It is likely that the same outcome would have taken place regardless of a composite being created or released.

 

[MadMcGoo]

Just to back up my opinion, I looked at the posters pertaining to [ARRESTED] or [GUILTY] perps. I know that Parabon is just using those as an example of their successful AI technology, but imo it's a bit misleading because it almost implies that the creation of the snapshot composite LED to the arrest of that person (or I thought that's what it implied, lol). Clicking on the links and reading the articles, majority of those cases were solved through genetic genealogy or other means, either done through Parabon or another facility. It is likely that the same outcome would have taken place regardless of a composite being created or released.

I think the question here is did the sketch resemble the perpetrator or victim. Regardless of how the identification was made, were the Parabon recons accurate or not. JMHO

 

[bombardier]

I think the question here is did the sketch resemble the perpetrator or victim. Regardless of how the identification was made, were the Parabon recons accurate or not. JMHO

Honestly the comps of suspects that Parabon shares are mostly accurate (when considering hairstyles/BMI/etc differences) but another question is...even if generically accurate, if they aren't themselves leading to the arrests, what is the benefit of releasing these images? If there are no 'pros', other than maybe getting some more MSM coverage for certain cases, are there actually 'cons' that come into play that should sway ones decision to release such images?
On post #1, social media users are saying that the generic image of the young black perp may lead to innocent look-alikes being targeted or harassed. I would be interested to see if this has any backing, and unnecessary harassment has ever come from a phenotype composite.

 

[MadMcGoo]

Honestly the comps of suspects that Parabon shares are mostly accurate (when considering hairstyles/BMI/etc differences) but another question is...even if generically accurate, if they aren't themselves leading to the arrests, what is the benefit of releasing these images? If there are no 'pros', other than maybe getting some more MSM coverage for certain cases, are there actually 'cons' that come into play that should sway ones decision to release such images?
On post #1, social media users are saying that the generic image of the young black perp may lead to innocent look-alikes being targeted or harassed. I would be interested to see if this has any backing, and unnecessary harassment has ever come from a phenotype composite.

Sure, more MSM coverage helps. BUT, as with any case, I’m positive the release of these images brings in dozens of unrelated tips. Which in turn probably results in harassment of people not involved.

Hindsight is always 20/20, so we can ascertain the results of the composites release afterwards and wether or not it played a part in the identification. It would be interesting to know statistics on this!

 

[MadMcGoo]

I guess, for me, this is no different than releasing a pencil sketch of a dark-haired white male driving a VW Bug in the 1970s to newspapers.

 

[CCJD]

what is the benefit of releasing these images? If there are no 'pros', other than maybe getting some more MSM coverage for certain cases, are there actually 'cons' that come into play that should sway ones decision to release such images?

rsbm
Here's one where a snapshot could be beneficial - Paradise John Doe. No head recovered.

Kind of funny/odd that NamUs doesn't have it on his case.

The National Missing and Unidentified Persons System (NamUs)

By bringing people, information, forensic science and technology together, NamUs helps resolve cases.

www.namus.gov

but PA Crime Stoppers does

CASE: COLD CASE - Unidentified Male Torso - Paradise Township - Monroe County - $5000

DATE FOUND: December 26, 2011LOCATION: Area off of SR 191, approximately 6/10 mile south of Sylvan Cascade Road in Paradise Township, Monroe County.CIRCUMSTANCES: A citizen walking along SR 191 in Paradise Township, Monroe County discovered the

www.crimewatchpa.com

link for photo: Police Predictions - Parabon® Snapshot® DNA Analysis Service

 

[bombardier]

rsbm
Here's one where a snapshot could be beneficial - Paradise John Doe. No head recovered.

Kind of funny/odd that NamUs doesn't have it on his case.

The National Missing and Unidentified Persons System (NamUs)

By bringing people, information, forensic science and technology together, NamUs helps resolve cases.

www.namus.gov

but PA Crime Stoppers does

CASE: COLD CASE - Unidentified Male Torso - Paradise Township - Monroe County - $5000

DATE FOUND: December 26, 2011LOCATION: Area off of SR 191, approximately 6/10 mile south of Sylvan Cascade Road in Paradise Township, Monroe County.CIRCUMSTANCES: A citizen walking along SR 191 in Paradise Township, Monroe County discovered the

www.crimewatchpa.com

link for photo: Police Predictions - Parabon® Snapshot® DNA Analysis Service

Oh don't get me wrong, I think it's a great tool for John and Jane Doe cases!! Especially when theres extra info (clothing, circumstances, hair found w remains) to add to and specialize the composite.

 

[ntt1]

I don't think that releasing the composite in this case was particularly productive, but the angry Twitter mob responses I've seen I feel are oversimplifying and making a lot of absolute statements on things that definitely more complex.

These images are always released with caveats that it is an inexact science and they are not meant to be exact likenesses, but the reality is that a lot of people will take composite images at face value (heck, look at all the "it looks nothing like the person!" comments whenever someone is identified). Thus there are definitely legitimate concerns among certain communities who have historically been marginalized by the police. With that being said, I'm not sure spending the money on this is the best use of police resources when FGG is available, but to completely disregard it as bunk or pseudoscience to me seems a bit short sighted.

They have the perpetrator's DNA presumably a sexual assault kit, which would presumably be used for comparison with a suspect, so I am not sure where the wrongful conviction concerns come from here.