Accused murderer wins right to check source code of DNA testing kit used by police
A New Jersey appeals court has ruled that a man accused of murder is entitled to review proprietary genetic testing software to challenge evidence presented against him. Attorneys defending Corey Pickett, on trial for a fatal Jersey City shooting that occurred in 2017, have been trying to examine the source code of a software program called TrueAllele to assess its reliability. The software helped analyze a genetic sample from a weapon that was used to tie the defendant to the crime. The maker of the software, Cybergenetics, has insisted in lower court proceedings that the program's source code is a trade secret. The co-founder of the company, Mark Perlin, is said to have argued against source code analysis by claiming that the program, consisting of 170,000 lines of MATLAB code, is so dense it would take eight and a half years to review at a rate of ten lines an hour. The company offered the defense access under tightly controlled conditions outlined in a non-disclosure agreement, which included accepting a $1m liability fine in the event code details leaked. But the defense team objected to the conditions, which they argued would hinder their evaluation and would deter any expert witness from participating. With the two sides unable to reach an agreement, the lower court judge hearing the case denied the defense's motion to vet the source code, without sufficiently addressing the defense arguments. That prompted the defense to appeal. Those arguing on behalf of the defense cited past problems with other genetic testing software such as STRmix and FST (Forensic Statistical Tool). Defense expert witnesses Mats Heimdahl and Jeanna Matthews, for example, said that STRmix had 13 coding errors that affected 60 criminal cases, errors not revealed until a source code review. They also pointed out, as the appeals court ruling describes, how an FST source code review "uncovered that a 'secret function ... was present in the software, tending to overestimate the likelihood of guilt.'" On Wednesday, the appellate court sided with the defense [PDF] and sent the case back to a lower court directing the judge to compel Cybergenetics to make the TrueAllele code available to the defense team. "Without scrutinizing its software's source code – a human-made set of instructions that may contain bugs, glitches, and defects – in the context of an adversarial system, no finding that it properly implements the underlying science could realistically be made," the ruling says. Kit Walsh, senior staff attorney for the Electronic Frontier Foundation, hailed the appellate ruling. "No one should be imprisoned or executed based on secret evidence that cannot be fairly evaluated for its reliability, and the ruling in this case will help prevent that injustice," she said in a blog post. If TrueAllele is found wanting, presumably that will not affect the dozen individuals said to have been exonerated by the software. Una buona notizia tratta da https://www.theregister.com/2021/02/04/dna_testing_software/ Giacomo
On Tue, Feb 09, 2021 15:35:37 PM +0100, Giacomo Tesio wrote:
A New Jersey appeals court has ruled that a man accused of murder is entitled to review proprietary genetic testing software to challenge evidence presented against him. ... Una buona notizia tratta da https://www.theregister.com/2021/02/04/dna_testing_software/
Buona notizia davvero, e molto interessante pure. Grazie! Marco -- Help me to continue my digital awareness blogging: https://stop.zona-m.net/christmas-2020-fundraising/
Grazie Giacomo! basta grattare appena appena la superficie per osservare che *pare* che un pochino di scienza e un pochino di tecnica sono... taroccate. «The answer my friend, is blowing in the wind, the answer is blowing in the wind». Giacomo Tesio <giacomo@tesio.it> writes: [...]
Those arguing on behalf of the defense cited past problems with other genetic testing software such as STRmix and FST (Forensic Statistical Tool).
Il tema generale è quello della tecnica denominata «Probabilistic genotyping» https://en.wikipedia.org/wiki/Probabilistic_genotyping --8<---------------cut here---------------start------------->8--- Probabilistic genotyping is the use of statistical methods and mathematical algorithms in DNA Profiling. It may be used instead of manual methods in difficult situations, such as when a DNA sample is very small or includes a mixture of multiple individuals' DNA.[1] Probabilistic genotyping, unlike traditional methods, avoids the need for subjective judgment.[1] The reliability of the method has been questioned by some defense lawyers because the source code of some probabilistic genotyping programs is proprietary.[2] --8<---------------cut here---------------end--------------->8--- [1] https://jolt.law.harvard.edu/assets/articlePDFs/v31/31HarvJLTech275.pdf [2] http://www.abajournal.com/magazine/article/code_of_science_defense_lawyers_w...
Defense expert witnesses Mats Heimdahl and Jeanna Matthews, for example, said that STRmix had 13 coding errors that affected 60 criminal cases, errors not revealed until a source code review.
Software validato e approvato dal Department for Forensic Sciences USA: https://dfs.dc.gov/sites/default/files/dc/sites/dfs/page_content/attachments... ...però non trovo un documento che elenchi o spieghi i 13 coding errors trovati in STRmix
They also pointed out, as the appeals court ruling describes, how an FST source code review "uncovered that a 'secret function ... was present in the software, tending to overestimate the likelihood of guilt.'"
Per approfondire questo (FST) l'articolo stesso include il link a questo: «Federal Judge Unseals New York Crime Lab’s Software for Analyzing DNA Evidence» https://www.propublica.org/article/federal-judge-unseals-new-york-crime-labs... In quel caso il giudice ha ordinato la pubblicazione del sorgente. [...]
"Without scrutinizing its software's source code – a human-made set of instructions that may contain bugs, glitches, and defects – in the context of an adversarial system, no finding that it properly implements the underlying science could realistically be made," the ruling says.
Ecco perché sarebbe proprio il caso che fosse scritto a chiare lettere nei codici di procedura penale che il software utilizzato per raccogliere "evidenze scientifiche" valide nei processi **deve** essere software libero (e riproducibile). Non solo, sarebbe *anche* il caso di migliorare drasticamente la riproducibilità delle ricerche scientifiche, soprattutto quando usate per raccogliere "evidenze scientifiche" nei processi o per supportare decisioni chessò, di politica economica come quindici anni di austerity :-O ... ... perché: «Federal Court Says 'Touch DNA' Analysis Is Mostly Guesswork That Can't Be Used As Evidence» Oct 30th 2019 https://www.techdirt.com/articles/20191026/12493943266/federal-court-says-to... --8<---------------cut here---------------start------------->8--- [...] But the myth that DNA evidence is nigh-infallible persists. Some of this is due to the inscrutable nature of the processes that turn stray cells into evidence. Some of this is due to forensic experts overstating the certainty of their findings. When DNA evidence is pretty much the only evidence holding a case together, the evidence had better be solid. A federal court in Michigan has found that the framework behind one company's (STRmix) DNA evidence testing is a cobbled-together mess that sounds nice and science-y, but isn't much more than overly-educated guesswork. [...] The court points to the lack of standardization across law enforcement agencies utilizing the software as well as the lack of objective peer review of research written and published by the developers of STRmix. Throwing a bunch of science together in a blender, like so many DNA samples pulled from a doorknob, is no way to make definitive declarations about probability. --8<---------------cut here---------------end--------------->8--- L'articolo sopra contiene interessantissimi link, tra i quali: 1. «Study Buried For Four Years Shows Crime Lab DNA Testing Is Severely Flawed» Sep 27th 2018 https://www.techdirt.com/articles/20180924/18174740705/study-buried-four-yea... 2. «DOJ/FBI Admit They May Have Abused Hair Analysis To Convict Hundreds To Thousands Of Innocent People» Jul 23rd 2013 https://www.techdirt.com/articles/20130723/00563923895/dojfbi-admit-they-may... Ohibò! [...] Saluti, Giovanni. P.S.: https://en.wikipedia.org/wiki/Phantom_of_Heilbronn «The cotton swabs used by many state police departments were found to have been contaminated before shipping.» -- Giovanni Biscuolo
Ciao Giovanni, On February 10, 2021 6:32:01 PM UTC, Giovanni Biscuolo wrote:
il software utilizzato per raccogliere "evidenze scientifiche" valide nei processi **deve** essere software libero (e riproducibile).
In generale, se il software ha effetti di legge, usare software proprietario è come avere leggi segrete. Se poi l'algoritmo eseguito da quel software poi non è completamente deducibile dal suo codice, bug inclusi (perché ad esempio contiene una ANN), allora è come avere leggi inspiegabili ed incomprensibili by design. E per quanto ne so, Leggi segrete e/o inspiegabili non hanno posto in una democrazia. Giacomo
participants (3)
-
Giacomo Tesio -
Giovanni Biscuolo -
M. Fioretti