Caption: Ronnie Boodoosingh
- Human Element First: Speaking to the media in Port of Spain on Thursday, Chief Justice Ronnie Boodoosingh stressed that while artificial intelligence (AI) is a useful tool, judicial decision-making requires human emotion, reason, empathy and compassion.
- Permitted Uses & Verification: After reviewing nearly two dozen international models, the Judiciary developed a practice direction allowing AI for administrative tasks—such as summarising pleadings or generating chronologies—provided all outputs are strictly verified by a human.
- Consequences for Misuse: Attorneys who use AI outside the prescribed rules risk having their submissions struck out or facing contempt of court, while the policy remains subject to regular review as the technology evolves.
By Prior Beharry
While artificial intelligence (AI) can serve as a valuable tool to assist the courts, it can never replace the human qualities of emotion, reason, empathy and compassion required in judicial decision-making, Chief Justice Ronnie Boodoosingh said on Thursday.
Boodoosingh made the statement while responding to a question from AZP News during a media interview at the Judiciary’s training room at Tower D of the International Waterfront Centre in Port of Spain.
Explaining that as Chief Justice he cannot dictate how judges decide their cases, Boodoosingh noted that the Judiciary instead developed a formal policy and practice direction to guide how AI should be used.
To craft the framework, the Judiciary examined nearly two dozen practice directions from around the world and consulted with information technology (IT) and industry experts, he said.
Boodoosingh said, “What we came up with is a mechanism where this important tool—just like virtual hearings, just like our phones where we Google things and get information—AI is a means by which a tool can assist us.”
He said, “So we said in the actual decision-making, people do not come for a robot to make a decision for them. They come for a judicial officer who has emotion, who has the ability to reason, who has empathy, who has compassion, who has understanding of human nature. All of those factors, those are the factors in judicial decision-making that are absolutely critical for a human being to exercise.”
Striking a balance
Boodoosingh said that the Judiciary’s approach aims to strike a practical balance by using AI to speed up time-consuming administrative tasks while keeping human verification mandatory.
He said, “There are some tasks, for example, summarising of documents or pleadings and so on. It can be done, but you have to test it and verify it and make sure it is an accurate summary. It can give you a summary of a judgment previously given, but you have to read the judgment and make sure that that is an accurate summary of the judgment.”
He noted that tasks that previously took three hours, such as generating a chronology of important facts, can now be completed in seconds using AI, provided someone verifies that the facts are correct.
The Chief Justice added that the issue also formed part of his recent discussions with judges in the United Kingdom, who similarly maintained that AI will not be used in actual decision-making.
Consequences for misuse
Under the Judiciary’s practice direction, attorneys-at-law who misuse AI face strict penalties.
He said, “We have built in certain things in the practice direction. So if, in certain circumstances, it could be a contempt of court for an attorney-at-law to use AI in a way not prescribed, it might be a matter that the court judge will say, ‘I pay no regard to this application or these submissions, I’m going to strike them out.’”
When asked what consequences a judge would face for using AI improperly, Boodoosingh said judicial officers are expected to be guided by honour and the principles set by the Judiciary.
“Well, you see, we have guidance for judicial officers. So the idea of the judge using it wrongly is probably too strong a word,” he said, noting that action would likely require “some kind of blatant disregard” of the policy.
He said, “But there are avenues. It could begin with a pastoral conversation, that kind of thing. Well, one would not expect that our judicial officers would do that.”
Boodoosingh said that because AI technology is rapidly evolving, the practice direction remains subject to ongoing review—whether in six months or a year—with continued input from industry experts.
