Forensic Evidence

Election Security: US Lawmakers Concerned “Deepfake” Videos are the Next Stage of Information Warfare Ahead of 2020 Election

February 6, 2019
Elections, Ethics, Foreign Intelligence, Forensic Evidence

By: Jack Kall

The nation’s attention has turned to the 2020 election with the 2018 midterms in the rear view mirror. Accordingly, an increasing number of US lawmakers are concerned that a form of video manipulation known as “Deepfakes” will be the next stage of information warfare. In short, Deepfake videos are hyper-realistic manipulated videos made using artificial intelligence technology. The videos are often convincing enough that it can be difficult to even tell what has or has not been manipulated. To raise attention, BuzzFeed published this video of Barack Obama delivering a public service announcement regarding dangers of the technology—except it was actually Jordan Peele.

Election security is a more important issue for US voters in the wake of Russian-led election interference in the 2016 Presidential Election. A recent Pew Research poll found that 55% of Americans say they are not too (37%) or not at all (17%) confident that elections systems are secure from hacking and other technological threats. Republicans (59% at least somewhat confident in security) express greater confidence than Democrats (34%), which is a reversal of attitudes from 2016.

While the threat of deepfakes has not garnered the same attention as Russian interference and other forms of “Fake News,” some US legislators are beginning to vocalize concern. This past September, three members of the House of Representatives—including the new chair of the House Intelligence Committee Rep. Adam Schiff (D-CA)—sent a letter expressing concern that the “technology could soon be deployed by malicious foreign actors” to the Director of National Intelligence Dan Coates. Senator Marco Rubio (R-FL) also displayed concern for the technology at a Senate Intelligence Committee by describing a scenario in which a deepfake video is released just before an election and going viral before analysts could determine it was fake.

While concern is rising, there is still a shortage of solutions. In January 2019, House Democrats unveiled several election security measures, but lacked solutions for deepfakes. The same month, Brookings Institute released advice for campaigns to protect against deepfakes. It remains to be seen whether Brooking Institute’s advice to protect infrastructure, add two-factor authentication, film the candidate at speaking engagements, and replicate a classified environment—while important general advice—is enough to protect against this ever-evolving deepfake technology.

Chimeras in DNA Forensic Testing: What to do?

February 13, 2013
Criminal Law, DNA Evidence, Forensic Evidence, Forensics

by Ryan J. Connell, UMN Law Student, MJLST Staff

The answer as suggested in an essay titled Chimeric Criminals by David H. Kaye in the current issue of the Minnesota Journal of Law, Science and Technology is not to worry about it too much.

The article criticizes the book Genetic Justice: DNA Databanks, Criminal Investigations, and Civil Liberties by Sheldon Krimsky and Tania Simoncelli. The book has latched on to a particular genetic anomaly referred to as chimerism. Chimerism denotes the presence of two genetically distinct cell lines in the human body. The authors of Genetic Justice want to use this rare condition to show that the supposed assumption that DNA profiling is infallible is incorrect.

Think for a moment about what DNA evidence has done in criminal law. Do not just think of the convictions, but think of the acquittals, and think of those freed from incarceration by innocence projects around the country that can be attributed to the use of DNA evidence. To call DNA evidence into question over such a rare and insignificant condition such a chimerism stretches the confines of reasonableness. Genetic Justice proffers that there is a 1/2400, 1/10, 1/8, and 1/1 incidence of chimerism. Other estimates are no better. A 2010 article in the Globe and Mail entitled “The Dark Side of DNA” called DNA evidence into question and offered that chimerism may be present in anywhere from a tiny population to ten percent of the population. If an entire science is going to be called into question some better statistics might be advisable first.

This book and other sources, such as “Expert evidence: the genetic chimerism and its implications for the world of law” by Daniel Bezerra Bevenuto assume that if genetic evidence is gathered and then does not match the defendant’s DNA that the courts and lawyers will simply dismiss the case. I think courts can handle whatever problems chimerism presents. If DNA is recovered at a crime scene and identifies person X and said person is chimeric and the reference sample he provides doesn’t match the sample recovered at the crime scene the court will rightly be concerned. The natural and simple remedy to this solution is just to test again. Normally chimeric cells are isolated so a second reference sample taken from the suspect should resolve the anomaly.

Chimerism does not preset the problem that the authors of Genetic Justice suggest. It is a rare occurrence that a DNA sample recovered at a crime scene doesn’t match the DNA of the suspect it identifies. And even in those rare circumstances where the DNA doesn’t match it is an easy fix. For a more detailed analysis of this issue please read the article by David H. Kaye in the Minnesota Journal of Law Science and Technology.

The full issue of MJLST in which David Kaye’s article appears can be found here.

Forensic Evidence

Contact Information