The “deepfake paradox” could undermine the justice system

The mere existence of deepfake technology casts doubt on all video evidence, posing new challenges to the justice system.

Video evidence appeared in approximately 80% of criminal cases in 2016, and the number may be much higher today, due to the increased use of CCTV cameras and smartphones with high-resolution cameras.

However, the arrival of deepfake technology, which uses AI to generate highly convincing yet fake videos, images, and audio, may bring an end to the type of evidence once considered so compelling.

The “deepfake paradox” – the erosion of inherited trust in video evidence due to the existence of deepfake technology – challenges the justice system, which was built on the assumption that “seeing is believing,” according to AI researcher Chiara Gallesse.

“In five years, a defendant will stand in court and say ‘that's a deepfake,’ and the prosecution won't be able to prove otherwise beyond a reasonable doubt,” Gallesse wrote in a LinkedIn post.

The US Judicial Conference’s Advisory Committee decided in May that an amendment to Rule 901 to create a specialized authentication process for potential deepfakes was not necessary because so few deepfakes had been offered as evidence at the time.

However, lawyers tell Cybernews that the deepfake technology is already finding its way into courts, and may have far-reaching implications for the judicial system.

Deepfakes may boost the “reverse CSI effect”

Deepfakes could erode the public’s trust in audio and video evidence, which has long been a cornerstone of both civil and criminal cases, says David Dozier, an attorney and managing partner at Dozier Law.

He has already seen instances where a defendant sought to raise doubt about legitimate video evidence by questioning its authenticity without presenting any evidence in support of the argument.

As a result, the courts may need to develop more standardized authentication protocols.

Dozier tells Cybernews, “One idea that I’ve seen starting to gain traction in legal circles is the notion of requiring 'origin certificates' for recorded digital content – embedded attestations that can be traced all the way back to the device from which they were captured.”

The existence of deepfake technology may contribute to the “reverse CSI effect,” a term coined after a popular forensic TV show, which refers to unrealistic jurors’ expectations of being provided with high-quality forensic evidence, Riana Pfefferkorn, a researcher at Stanford Law School, argues in her paper.

In other words, jurors may start expecting the proponent of a video to use sophisticated technology to prove to their satisfaction that the video is not fake

“If juries – entrusted with the crucial role of finders of fact – start to doubt that it is possible to know what is real, their skepticism could undermine the justice system as a whole,” the paper reads.

One idea that I’ve seen starting to gain traction in legal circles is the notion of requiring 'origin certificates' for recorded digital content – embedded attestations that can be traced all the way back to the device from which they were captured.

David Dozier

Courts will probably hear more “liar’s dividend” claims, such as someone denying evidence just because the deepfake technology is there, says Thomas Fighter, a founder and criminal trial attorney at Fighter Law.

As a trial lawyer, he already advises clients to obtain electronic evidence early, preserve the original, and not transmit recordings through multiple devices.

The conversation on this topic is live. Join in the discussion.

Fighter says he is most concerned with equity – defendants must be able to contest evidence, but that victims, too, must have credible means to establish the truth.

“Deepfake technology upsets that balance, but it also supports a fundamental constitutional principle – justice demands both accuracy and accountability, and the legal system will always adapt to defend that,” he says.

The system is bracing itself for more deepfakes

At the same time, deepfake technology is used to provide forged evidence, adding a new level of fear to trial attorneys’ minds because of how realistic it can seem, says James Rubinowitz, a founder and attorney at Rubinowitz Law Firm.

​However, it isn’t an entirely new problem, as the courts have successfully dealt with bad actors who tried to introduce doctored photographs into evidence for years.

​The court system requires authentication of photographic and video evidence before it can be shown to a jury.

​“This can be done either by a witness with personal knowledge of the incident who can verify the evidence is real, or by a technology expert who can establish an unadulterated chain of custody,” Rubinowitz explains.

A California court dismissed a case over AI-doctored evidence in September after a judge was able to identify that two Ring cam videos presented by a plaintiff were fake. The giveaways included a lack of facial expressions and the looping video feed.

David Fish, a partner at Romano Law, says the California case isn’t an isolated incident, and there are more to come.

“If key players in the legal system are not aware of the signs of deep-fake media, it could lead to incorrect rulings and injustice,” he adds.

However, Pfefferkorn predicts the panic around deepfakes in courts will prove largely unfounded. The introduction of image editing software Adobe Photoshop in the 1990s was also accompanied by warnings of a “crisis of truth,” but society adapted to the technology.

She writes, “The nation’s courts are robust institutions that have shown themselves capable of handling each new variant of the age-old problem of fakery.”

Unlock more exclusive Cybernews content on YouTube