Utah courts rely on long-standing evidence rules, but those rules now meet new challenges when AI enters the picture. The biggest issues involve:

• Whether AI-generated content can be authenticated
• Whether metadata and chain-of-custody records are intact
• Whether the content violates privacy or impersonation statutes
• Whether expert testimony is required to explain how the content was created

In Utah, the party introducing a video, audio clip, image, or document must show it is what they claim it is. With deepfakes, that burden is harder to meet because synthetic media can be nearly indistinguishable from real recordings.

Utah uses Rule 901 of the Utah Rules of Evidence, which requires enough proof to show that a piece of evidence is authentic. For AI-altered material, courts often expect supporting proof such as:

Chain of custody matters even more with deepfakes because the risk of alteration is high. Utah digital-forensics teams typically examine:

• Timestamps and any available geolocation data
• File-creation and modification history across devices
• Hash values before and after transfers or downloads
• How the evidence was stored, preserved, and collected

A break in the chain of custody does not automatically exclude the evidence, but it gives defense attorneys an opening to challenge reliability. Judges weigh how serious the gaps are and whether there is enough other proof to trust what the file shows.

Deepfake technology is increasingly used for impersonation, harassment, and fraud. Even if a fake video never makes it into evidence, the act of creating or sharing it can trigger Utah criminal statutes, including:

• Criminal impersonation and identity-related offenses
• Identity fraud and computer-crime statutes
• Harassment, cyberstalking, or intimidation crimes
• Voyeurism or distribution-of-intimate-images laws in sensitive cases

A deepfake used to mimic a defendant, victim, or witness can lead to additional charges and can also undermine the legitimacy of evidence presented during a criminal case.

Utah judges frequently rely on digital-forensics experts when AI-generated media is disputed. Experts may testify about:

• Whether the file shows signs of manipulation
• Whether AI-based artifacts or inconsistencies are present
• How deepfake models typically distort frames, lighting, or audio
• What level of certainty exists regarding authenticity

Expert testimony helps courts apply Utah reliability standards without requiring judges or jurors to understand every technical detail about AI. The goal is to translate complex analysis into plain, credible explanations the court can use.

This chart summarizes the main review points Utah courts consider when deepfakes or other AI-generated files are offered as evidence:

Q: Are deepfake videos automatically inadmissible in Utah?
A: No. Utah courts evaluate them under traditional authentication standards. They must be shown to be reliable, properly sourced, and not misleading.

Q: Can AI-generated audio or screenshots be used as evidence?
A: Yes, but the party offering them must prove authenticity. For AI-generated materials, this often requires expert analysis and clear supporting proof.

Q: What if a defendant claims an incriminating video is a deepfake?
A: Utah courts may order forensic evaluation. Judges look for inconsistencies, metadata anomalies, and evidence that the file came from a trustworthy source.

Q: Can someone be charged for making a deepfake of another person?
A: Depending on intent and content, Utah criminal laws on impersonation, harassment, computer crime, or image-based abuse may apply.

Q: Can police use AI tools to create evidence?
A: AI tools can assist investigations, but law enforcement cannot fabricate evidence. Courts scrutinize any AI-generated material extremely closely.