TRINIDAD AND TOBAGO: Judiciary Issues Practice Direction On Responsible Use Of A.I. In Proceedings

Media Release: Judiciary Issues Practice Direction on the Responsible Use of Generative Artificial Intelligence in Court Proceedings (10.07.26)

Judiciary Issues Practice Direction on the Responsible Use of Generative Artificial Intelligence in Court Proceedings

JULY 10, 2026

Port of Spain, Trinidad and Tobago – July 10, 2026: The Judiciary of Trinidad and Tobago has issued a new Practice Direction on the Use of Generative Artificial Intelligence (Gen AI), establishing a framework for the responsible and ethical use of AI in court proceedings.

Issued by Chief Justice Ronnie Boodoosingh on July 7th 2026, the Practice Direction took
effect on July 8th, 2026, and was published in the Trinidad and Tobago Gazette on July 9th, 2026. It provides guidance for judicial officers, attorneys-at-law, litigants, witnesses, experts and other court users on the appropriate use of generative AI within the justice system.

The Practice Direction supports the Judiciary's ongoing efforts to modernise the administration of justice while ensuring that judicial decision-making remains grounded in human judgment, professional ethics, fairness and accountability.

Recognising both the opportunities and risks presented by rapidly evolving AI technologies, the Practice Direction addresses concerns such as fabricated legal authorities, inaccurate citations, bias, confidentiality risks and the misuse of AI-generated content in court filings.

Among its key provisions, the Practice Direction:

- affirms that court proceedings must remain human-controlled and that AI cannot replace independent legal or judicial judgment;
- permits the limited use of AI for drafting and administrative purposes, subject to verification of all facts, authorities and legal citations;
- prohibits the use of AI to generate evidence, including affidavits and witness statements, except for limited formatting and language assistance;
- requires disclosure whenever AI has been used in preparing court documents, including certification of the AI tool used and confirmation that all legal authorities have been independently verified;

- restricts the use of AI by judicial officers to administrative and research functions and prohibits its use in judicial reasoning or drafting judgments; and
- empowers the Court to require additional disclosure and impose sanctions where AI is misused in court proceedings.

Recently commenting on the role of artificial intelligence in the justice system, Chief Justice Ronnie Boodoosingh stated:

"We should not wish AI to be deciding cases for us... It can provide useful data for us to make human decisions."

He further noted:

"The question is not whether AI should be used, but how can we responsibly use AI as a tool to assist us in making the human judgments that we need to make?"

The full Practice Direction on the Use of Generative Artificial Intelligence in Court Proceedings is available on the Judiciary of Trinidad and Tobago's website and in the Trinidad and Tobago Gazette. Members of the legal profession, court users and the public are encouraged to familiarise themselves with the new requirements.

####

Summary: Gen AI Use Guidelines in Legal Proceedings (Trinidad and Tobago)

Reasoning and Approach

To provide a clear and actionable summary, I have extracted the main points from the official Practice Direction on the use of Generative Artificial Intelligence (Gen AI) in legal proceedings in Trinidad and Tobago, as published in the Gazette No. 118 of 2026. The summary below is structured by user type and document type, highlighting key rules, risks, and required practices. Examples are included for clarity.

1. General Principles

• Human Control: Court proceedings must remain human-driven. Gen AI is a tool, not a substitute for professional judgment.
• Responsible Use: Use Gen AI in a way that upholds ethics, fairness, and independent judgment.
• Risks to Consider:
  • Potential for false or erroneous outputs (e.g., fake cases, incorrect citations).
  • Data quality and bias issues.
  • Confidentiality and privacy risks, especially with open-source tools.
  • Possible copyright breaches in Gen AI training data.

2. Guidelines by Document/User Type

A. Pleadings and Applications

• Gen AI may be used to draft pleadings/applications, similar to using precedent books.
• Obligation: Parties must verify all facts before signing.
• Example: Using Gen AI to draft a claim form, but manually checking all facts before submission.

B. Affidavits and Witness Statements

• Prohibited: Gen AI must NOT generate the content of affidavits or witness statements.
• Permitted: Formatting, spell-check, and grammar-check only.
• Reason: These documents must reflect the witness’s own words and knowledge.
• Example: Using Gen AI to correct grammar in a witness statement, but not to write the statement itself.

C. Submissions (Summaries, Chronologies, Notes)

• Gen AI may assist in preparing summaries, chronologies, oral/written submissions, and speaking notes.
• Verification Required: All citations, quotations, and legal authorities must be checked for existence, accuracy, and relevance.
• Responsibility: The court user is fully responsible for the content.
• Example: Using Gen AI to draft a summary of facts, then verifying every case and statute cited.

D. Expert Reports

• Gen AI must NOT draft or prepare the main content of expert reports.
• Permitted: Summarising content, generating tables, visual aids, indexes, and chronologies.
• Expert’s Duty: Verify all AI-generated content.
• Example: Using Gen AI to create a timeline chart for an expert report, but not to write the expert’s opinion.

E. Other Uses

• Gen AI can generate chronologies, indexes, and summarise/review documents or transcripts.
• Verification: Accuracy must always be checked.
• Confidentiality: Do not input sensitive or privileged information into open-source Gen AI tools.

F. Certification Requirement

• If Gen AI is used in preparing a court document, include a certificate at the end:

  Quote

  “This document was prepared with the assistance of Gen AI (state Tool and Version) as follows: (briefly describe, in less than 100 words, how it was used). All legal authorities cited have been verified by the author(s).”

G. Judicial Officers

• May use Gen AI for summaries of fact, evidence, chronologies, and issue identification.
• Prohibited: Using Gen AI for drafting reasons/rationale for judgments or for editing/proofing draft judgments.
• Disclosure: Judicial Research Counsel must disclose Gen AI use to the judicial officer.
• Responsibility: Judgments must remain the work of the judicial officer.
• Sanctions: Misuse may result in striking out documents, disciplinary action, or contempt of court.

3. Training and Review

• Ongoing training for judicial officers and court users is encouraged.
• Guidelines will be periodically reviewed due to rapid AI developments.

In summary: Gen AI is permitted as a supportive tool for certain legal tasks, but not for generating core factual or evidentiary content. All outputs must be verified, and the ultimate responsibility for accuracy and integrity remains with the human user. Certification of Gen AI use is mandatory in court documents.

-END-

YOU CAN DOWNLOAD THE PDF DOCUMENT BELOW

JUDICIARY DIRECTION ON USE OF AI IN PROCEEDINGS.pdf