St. John’s Antigua – The Editor pressed her palm to her forehead and let out a deep sigh. It was the third time this week she had to rewrite a story submitted by one of her junior reporters. But this wasn’t just another case of sloppy grammar or an overlooked lead. The piece, meant to be a hard-hitting exposé on land rights in Barbuda, had been cobbled together by artificial intelligence.

Worse still, the AI got its islands confused – mixing up Barbuda with Barbados – and at one point quoted a U.S. policy document instead of a Caribbean one. The Editor was exasperated. “I want to tear my hair out,” she muttered under her breath, staring at a screen full of content that read more like a Wikipedia entry than a regional investigation. Quotes were restructured, the tone sterile, and the cultural context had vanished entirely. The journalist hadn’t fact-checked a single paragraph.

It’s a growing dilemma in Caribbean newsrooms. AI-generated stories are becoming more common – often mistaken for shortcuts to productivity – but when left unchecked, they erode editorial credibility and miss the nuance that gives regional journalism its soul.

In another corner of the island, a hardware store owner was grappling with a different kind of AI failure. The owner had hoped to modernise customer service. Instead, he lost a weekend’s worth of sales and had to apologise publicly on social media.

These are examples of an increasingly common dilemma across Caribbean newsrooms and businesses. Artificial intelligence, once hailed as a game-changing productivity tool, is now raising serious concerns – from eroding editorial integrity to exposing businesses to unprecedented cybersecurity threats.

What’s at risk

On average an Artificial Intelligence (AI) tool like Chat GPT and Claude which are more popularly used in the Caribbean, take about 20 – 60 seconds to generate an article, complete with headlines, and helpful tips written by algorithms and breached by invisible code.

They look impressive at first glance but if you aren’t careful, publishing without fact-checking could mean defamation lawsuits or accusations of plagiarism. So, here’s where Caribbean journalists and business leaders find themselves at a crucial intersection.

On World Press Freedom Day this year, the Media Institute of the Caribbean (MIC) delivered a pointed warning. AI, though powerful, could pose an “existential threat to journalism in the region.”

Kiran Maharaj, MIC’s president, didn’t mince words. “AI could democratize information access, but without guardrails, it may erode the financial sustainability of Caribbean media,” she said.

Describing AI as a “double-edged sword,” Maharaj pointed to the promise it holds in streamlining verification and data analysis—but also to the risk it carries in spreading bias, misinformation, and detachment from regional realities. The algorithms, she noted, were designed “in and for Global North contexts.” They struggle with patois, local idioms, and even basic comprehension of our political history.

MIC Vice President Wesley Gibbings added a note of urgency: the region must “act swiftly” to educate journalists, adopt AI responsibly, and create policy frameworks to maintain public trust. Without that, he warned, “we risk becoming a region of recycled content, drowning in a sea of algorithmic noise.”

That message carried over to the Caribbean Media Summit in Guyana, where media professionals continued the conversation under the theme: “Enhancing Business Models in Turbulent Times.”

A Parallel Threat in the Business World

Just a day before the MIC’s call to action, business leaders and cybersecurity experts gathered at the American University of Antigua (AUA) for a conference hosted by C&W Business and the Antigua and Barbuda Chamber of Commerce. Their verdict? The biggest threat facing Caribbean enterprises today is unguarded data.

“Cyberattacks globally have risen by 47% in the first half of 2025,” warned Christian Benjamin of C&W Business, “but in the Latin American and Caribbean region, the figure is a staggering 108%.” Organisations in our region are now experiencing an average of 640 attacks per week.

AI is both accelerating the threats and widening the training gaps. CEO of Simply Secure, Kevin Gordon, said: “That for us is the biggest risk. It’s the biggest change in operation, it’ s also the biggest risk that we see.”

The myth that smaller Caribbean businesses are “too small to hack” has lulled many leaders into complacency. “So many think because they don’t see attacks, they’re safe. Wrong,” Gordon debunked.

According to Christian Benjamin, Associate Manager of Solution Architect Security/Networking, C&W Business, 99% of breaches occur at a management level. “They are busy with the volumes of emails they’re getting and they just click on anything. So, it has to start from the top.”

Kamla Hamilton, Senior Product Manager of Security at C&W Business drew attention to human error as a leading cause of cyber incidents, revealing that nearly 20% of attacks are the result of human interaction. “We are the ones that are taking on these ransomware, these phishing attempts,” she said, advocating for the development of a “human firewall” through staff training and cyber awareness.

Building this “human capital” must start at the top, panelists agreed, emphasizing that executives need to treat cybersecurity with the same seriousness they afford financial records and performance metrics.

How to make it work from a Caribbean Perspective

What connects these seemingly different warnings – from the newsroom to the boardroom – is the urgent need for Caribbean leadership to move from passive adoption to proactive governance. Despite the risks, AI still holds promise for Caribbean development if wielded with care.

For journalists that means learning how to properly use AI I tools and when to use them, and understanding legal issues.

For journalists that means learning how to properly use AI I tools and when to use them, and understanding legal issues. A recently published finding by the Media Institute of the Caribbean (MIC) titled AI & The Media Findings from the Caribbean Summit 2024 explores the impact of AI from a Caribbean perspective.

Here’s another helpful article published by the International Journalist network: ttps://ijnet.org/en/story/unlocking-power-genai-smarter-prompting-techniques

Both MIC and the Chamber’s conference pointed to policy gaps and outdated protocols as vulnerabilities. MIC has called for the creation of a Caribbean AI Ethics Task Force to help audit algorithms, combat bias, and promote transparency. 

Meanwhile, cybersecurity experts urged businesses to update systems – some haven’t revised their protocols in over a decade and begin treating digital threats with the same urgency as fire drills.

One proposed solution from MIC is particularly bold: taxing Big Tech and reinvesting proceeds into Caribbean journalism. It’s a controversial idea but one grounded in the reality that global platforms profit immensely from regional content and users without reinvesting proportionately.

No Matter what, Education is still Key!

Apart from your regular AI tools, generative AI can also -produce deep fakes, bot postings, and AI hallucinations which are all related to the spread of misinformation, but they involve different mechanisms. 

Deep fakes are AI-generated media, often videos, that convincingly mimic real people, potentially for deceptive purposes. 

Bot postings refer to automated social media accounts that disseminate content, which can include misinformation or propaganda.

AI hallucinations are errors or fabricated information generated by AI models, appearing as factual statements that are not true. 

So, it goes without saying that AI education should be a priority. Leading the way is the University of the West Indies Five Islands Campus in Antigua (UWI FIC) that is now currently hosting its 2nd annual UWI FIC Annual Research Conference.  

Antigua & Barbuda’s Prime Minister Gaston Browne also spoke to the possible practical real-world uses of AI in sharing information on drugs with manufacturers or creating our own drugs using AI, as well as using artificial intelligence to analyse CIP applications at a faster rate. 

“If we can get an AI system implemented at the CI unit, it can help to speed up the process,” Browne said. “You will still need to have some manual intervention to make the final decision. But… the system can actually assess the risk and indicate that this is an acceptable risk.”

Browne challenged students and scholars, particularly those studying on government scholarships, to consider how their research could help improve public sector operations. 

He added that an intelligent system could raise the integrity of the programme by identifying manipulation attempts that human reviewers might miss. “What the system would have said otherwise, when you would have done your checks,” he noted. Similar applications, he argued, could be used to bolster compliance in banking and improve government efficiency.