AI-Affected Evidence: Terms to Know

  • Unrealistically expecting conclusive scientific proof of guilt before the factfinder will convict, resulting in acquittals of guilty defendants because of lack of forensic evidence.

  • Placing too much weight on forensic evidence produced by prosecution, resulting in convictions in cases where the defendant probably should have been acquitted.

  • Whereby parties levy attacks on authentic materials that a fact finder might accept.

  • An assertion that all evidence is potentially synthetic and must be interrogated.

  • The propensity for jurors/fact finders to place disproportionate weight on visual information, even if instructed that it may be fake.

  • The inexplicability of certain AI systems, frustrating the auditability of an AI-affected process.