Last Updated on Friday, 18 April 2025, 21:26 by Writer
Guyana can use artificial intelligence (AI) as a tool to fight against misinformation and disinformation, a leading Caribbean expert on AI said Thursday.
“You can have AI agents that are actually scouring the Internet in a multilingual way – every language possible – for any hint of anyone that’s likely to mislead the Guyanese into thinking that something was about to happen and they can red flag and they can make recommendations about what to broadcast, when to broadcast and this can be done in an energetic way as well,” Dr Toussant Boyce, Head of the Office of Integrity, Compliance and Accountability at the Barbados-headquartered Caribbean Development Bank (CDB), told a lecture organised by the Guyana Press Association to mark next month’s annual observance of World Press Freedom Day.
He said that AI could assist in messaging in a way that is seamless and leverages technology including for those in far-flung areas. Dr Boyce also suggested that Guyana move away from legacy mass communication methods such as radio broadcasting to propagate its message. “The fact that we have to place on the radio and hope that certain persons hear this message is just an ‘interesting’ way to do it. Technology will be leveraged in the future to better deliver that precisely and effectively to those who need to use it most,” said Dr Boyce, a Caribbean-trained lawyer.
Dr Boyce, who has distinguished himself in the field of AI, said that aspect of technology could aid journalists in deciding whether or what types of assignments to cover.
He cautioned that deepfakes could confuse the media and public during elections.
On the point of elections he said Guyanese should be wary of the fact that AI could be used to create deceptive content in the names of now deceased politicians. “I don’t need to name names but you can expect, possibly, in Guyana’s context, as well, the extent to which that can be used,” he said.
Dr Boyce said nothing was stopping Caribbean nations from enacting legislation on AI.
He said the law could include safeguards for data that would be used in training, inferencing and access to data before it’s too late.
He said once a language model accesses information in journalism repositories, it would be virtually impossible to regain control.
