A Washington state judge blocked the use of AI-enhanced video as evidence in the first ruling of its kind

WhatsApp Group Join Now
Telegram Group Join Now
Instagram Group Join Now

A Washington state judge overseeing a triple-murder trial has banned the use of artificial intelligence-enhanced video as evidence in a ruling that experts say is the first of a United States criminal court It is a first of its kind event.

In a ruling signed Friday by King County Superior Court Judge Leroy McCullough and first reported by NBC News, the technology was described as novel and said it ‘thinks’ of an AI model. relies on fuzzy methods to represent

“This court finds that admitting this enhanced evidence would confuse the issues and disrupt the testimony of eyewitnesses, and the non-peer-review process used by the trial court.” A time spent within the .AI model could lead to a trial,” the judge wrote in a ruling that was posted in the docket Monday.

The decision comes as artificial intelligence and its uses β€” including the proliferation of deepfakes in social media and political campaigns β€” rapidly evolve, and state and federal lawmakers grapple with potential threats from the technology.

Attorneys for a man who opened fire outside a Seattle-area bar in 2021, killing three people and wounding two others, sought to introduce cellphone video through machine learning software, court filings show. . Machine learning is a special field within artificial intelligence that has emerged in recent years as the foundation of most modern AI systems.

According to a February filing in King County Superior Court, prosecutors in the case said there was no legal precedent for allowing the technology in a U.S. criminal court. Jonathan Hawke, a former lawyer and barrister in Canada and expert on photographic evidence in the United States and elsewhere, said it was the first case he knew of where a criminal court had considered the issue.

The defendant, Joshua Paluka, 46, has claimed self-defense in the Sept. 26 slaying, with his lawyers saying in a February court filing that he was attacked and shot. He was trying to reduce the situation of violence.

Plucka returned fire, killing innocent bystanders, the filing said. The man accused of attacking Paluka was also shot and killed, the probable cause statement showed.

The deadly collision was caught on cell phone video. To enhance the video, Pluka’s attorneys turned to a man who had never handled a criminal case before but had a background in creative video production and editing, according to prosecutors’ filings.

The software he used, developed by Texas-based Topaz Labs, says its software is used by film studios and other creative professionals to “supercharge” video, according to the filing.

Attorneys for Paluka did not respond to requests for comment. In a statement, a Topaz Labs spokesperson said the company “strongly” recommends using its AI technology for forensic or legal applications.

The prosecutor’s office said the enhanced video predicted images rather than reflecting the size, shape, edges and color captured in the original video. The enhanced images were “false, misleading and unreliable,” the filing said.

In a statement to prosecutors included in the filing, a forensic video analyst who reviewed the original and enhanced recordings said the enhanced version contained visual data that was not in the original. According to the expert, Grant Fredericks, data was also removed from the enhanced version.

Every pixel in AI-generated video is β€œnew, resulting in video that looks more pleasing to the eye of the average observer, but with the illusion of clarity and increased image resolution. which does not accurately represent the actual events. scene,” Fredericks wrote in the announcement.

In a separate filing, Plucka’s lawyers countered that such claims were “exaggerated and overblown.” A comparison of the two videos shows that the improved version is a “faithful depiction of the original,” the filing said. “And that’s what matters.”

In his statement, Fredericks, who has taught for the FBI and worked as a video analyst for 30 years, said he was unaware of peer-reviewed publications supporting AI video enhancements. Establish an accepted procedure. The FBI does not include anything on the subject in its best practices for handling forensic video, he said.

George Reiss, a former crime scene investigator and longtime forensic video analyst in Southern California, said he is aware of a handful of examples of artificial intelligence being used as a potential investigative tool to decipher license plate images. are

One of the companies that has developed such software, Amped, said in a post in February that artificial intelligence is not reliable enough to be used for image enhancement in a legal environment. The company pointed to the technology’s ambiguous results and potentially biased results.

“It’s a new science,” Reiss said. “It should be researched before someone uses it in real case work. It should be peer-reviewed. I’m not sure what level of AI will be appropriate for use at some point in the future.” I would actually explain the still photo or video, but it’s premature at this particular point.”

WhatsApp Group Join Now
Telegram Group Join Now
Instagram Group Join Now

Leave a Comment