Suggested A.I. Rule – Proposed Amendment to Maryland’s Computer-Generated Evidence Rule

Why It Matters

By formalizing procedural safeguards for AI evidence, the amendment protects trial integrity and reduces the risk of undisclosed or unreliable generative content influencing verdicts, setting a precedent for other jurisdictions.

Key Takeaways

• Amendment expands rule to include generative AI exhibits
• Requires 90‑day notice and full disclosure of AI data
• Allows pre‑trial hearing with expert appointment for objections
• Preserves AI evidence for appellate review
• Targets acknowledged AI; undisclosed deepfakes remain challenging

Pulse Analysis

The rapid integration of artificial intelligence into litigation has outpaced existing evidentiary rules, leaving courts to grapple with novel challenges such as hallucinated citations and AI‑generated visualizations. Maryland’s computer‑generated evidence rule, originally enacted in 1998 and last amended in 2009, has seen little use, yet its procedural structure provides a solid foundation for addressing modern AI outputs. By updating Rule 2‑504.3 to define generative AI evidence explicitly, the state acknowledges the technology’s growing role while preserving the rule’s focus on discovery, notice, and pre‑trial adjudication.

The proposed amendment introduces concrete deadlines—notice at least 90 days before trial and a 21‑day objection window—mirroring best‑practice recommendations from the Sedona Conference and Maryland’s own facial‑recognition statute. It also mandates disclosure of the AI model’s training data and methodology, enabling parties to assess reliability under Daubert‑type standards. A pre‑trial hearing, with the option to appoint an expert, ensures that judges receive technical guidance before ruling on admissibility, reducing the likelihood of surprise evidence derailing proceedings. Preservation requirements guarantee that AI exhibits are available for appellate scrutiny, reinforcing the rule’s appellate‑friendly design.

For litigators, the amendment offers a clear roadmap for introducing AI‑generated exhibits, turning what was previously a gray area into a predictable process. While it effectively curbs the use of “acknowledged” AI, undisclosed deepfakes remain a detection challenge, underscoring the need for complementary forensic tools and ethical guidelines. Maryland’s proactive stance may influence other states to adopt similar procedural safeguards, fostering a more consistent national approach to AI evidence and strengthening overall judicial reliability.