US Court Rules Against Use Of AI In Judicial Defense

A judge in the US state of Washington ruled against the use of artificial intelligence (AI) enhanced video evidence in a triple murder case due to the "novelty" and "opacity" of the technology.

The judge criticized AI technology as "novel" and "opaque," and said Monday it made it "challenging" to understand decisions regarding AI-enhanced video content.

The move is considered "unprecedented" in US criminal courts and stemmed from a case where the defense sought to introduce mobile phone footage, enhanced by machine-learning software, in the defense related to a shooting outside a bar near Seattle.

Prosecutors opposed the use of the technology, citing a "lack" of legal precedent, marking a legal development on AI-enhanced evidence admissibility.

The defense used machine learning, a type of AI, to improve mobile phone footage to support the defendant's claim of self-defense.

But Topaz Labs, developer of the software used by the defense, cautioned against using its technology for forensic or legal purposes, emphasizing its unsuitability.

An analysis made by Frederick Grant, a forensic video analyst, revealed disparities between original and enhanced videos, with other experts noting inaccuracies and misleading elements in the AI-enhanced version.

The use of AI may result in a video that may appear more "pleasing" but the illusion of clarity and increased image resolution may not "accurately" represent the events of the original scene, said Grant.

Despite defense assertions regarding its "fidelity," experts raised concerns about the lack of established methodologies and research supporting AI video enhancement.

While AI has seen limited exploration for investigative purposes such as to clarify license plate images, George Reis, a former crime scene investigator and longtime forensic video analyst stressed the necessity of rigorous research and standards before widespread use in legal settings.

"I'm not certain what level is going to be appropriate at some time in the future for the use of AI in actually doing a clarification of a still photograph or a video, but at this particular point it's premature," he said. -