Jamaican lawmakers have expressed concerns that the use of deepfakes and artificial intelligence (AI)-generated content will undermine the integrity of the democratic process and are agreed that there needs to be consensus on how the nation handles this growing threat.

The matter has taken on greater significance in the run-up to the local government election scheduled for February 26 and the fact that Jamaicans are due to vote again in parliamentary elections in 2025.

Deepfakes basically involve the use of learning algorithms, particularly generative adversarial networks, which can analyse and mimic patterns in data. Experts explain that they are used to alter images or voice recordings to misrepresent someone as doing or saying something that was not actually done or said.

Cabinet member Dr Christopher Tufton, who first pointed to the possible impact of deepfake on politics in Jamaica during a Jamaica Labour Party meeting in St Catherine last June, said he did so out of his personal monitoring of the global environment and how technology was being embraced for a number of vital services, including health, the portfolio for which he is responsible.

At the time, Tufton said that the use of artificial intelligence in political campaigning is posing a threat to democracy, and he urged the legislature to discuss the issue with a view to establishing sanctions for people using the technology to create mischief.

Last week, in an interview with the Jamaica Observer, Tufton reiterated the position, even as he acknowledged that “AI does have a lot of potential benefits”.

“At the same time, any assessment of social media space, TikTok in particular, you will see significant potential dangers of an abuse of the technology, and the persons at the highest level of the technology, like Elon Musk and so on, if you look at what their opinions are on it, you’ll hear those expert opinions around the need for regulation because there is a significant potential for abuse, and in the political sphere we have seen a lot of that,” Tufton said.

He pointed to the recent AI-generated nude images of American artiste Taylor Swift which went viral on the social platform X as an example of the abuse of the technology.

“So when I made the comment I saw it coming based on all that I had seen at the time. It has got more advanced in terms of usage and comfort levels, and I still remain concerned that if we do not establish some ground rules around engaging it, it can create major challenges for us and indeed damage people’s reputations, and create further levels of distrust within the society around how technology is utilised,” Tufton argued.

“I don’t want to take a posture to suggest that it is all bad, because there are many, many benefits of it; the question is how we regulate and manage it,” he told the Sunday Observer.

Tufton suggested that regulation is urgently needed as the problem “will only get more intense, either from the local government election but also a pending national election going forward over the next year and a half”.

“The political organisations need to take a joint approach around deciding on some ground rules, because otherwise it may become a free-for-all, and we have seen cases of it being used in a way that could be damaging and it threatens to compromise the entire democratic process, because people will approach things with a lot of suspicion; you have legal ramifications, reputations being destroyed,” he argued.

“You could have a case where something occurs on the eve of an election and by the time people validate or prove it false or otherwise the time would have passed, so you literally can use it, strategically based on timing, to disrupt people’s ambitions and interests and indeed the need for freedom of choice.

“So I believe it is something we need to look at — the possibility of the political organisations having a sort of joint agreement around the potential abuse of this technology,” Tufton insisted.

Julian Robinson, the Opposition spokesman on finance and Member of Parliament for St Andrew South Eastern, agreed and suggested that the Broadcasting Commission look at it with the view to starting a discussion on how best to manage the issue.

“It is something of great concern to me and to the political process in general, because to the unsuspecting public you can make up images, speeches of persons in the public domain and convey messages that are totally fake, and the challenge with it is the difficulty in responding and combating what is fake information,” Robinson told the
Sunday Observer.

“Things go viral in a very short period of time, and the person who is wronged will have to expend a lot of resources to correct what is put out there,” added Robinson.

He admitted that it is very difficult to regulate because anybody anywhere in the world can develop content. “You don’t have to be in Jamaica, so the question is, what is the most appropriate form of regulation in a scenario like this, and the resources that you would need to pin down the originator of something like this is also going to be significant,” Robinson said.

“You’re going to need people with great cyber-related skills to do it, so it is something that I believe as a country we would need to examine. I wouldn’t say I have the answers as to the best way to deal with it, because you always have to balance freedom of expression and the right to express yourself, versus where somebody is clearly putting out something that is fake news, and those things can have great consequences in an environment where news travels at lightning speed.”

Robinson, too, acknowledged that AI has a lot of potential and can be used to do a lot of great things. “But this is one of the areas where it has the potential to be damaging and can put anybody in a very negative light,” he said.

That view was shared by his fellow legislator Fitz Jackson, who represents St Catherine Southern in the Parliament. Jackson, though, was more strident on the need for legislation to deal with people using the technology for mischief.

“The technology without the appropriate regulations [and] safeguards can be very dangerous, and misrepresentation that can tarnish and damage persons’ reputation and well-being is something that ought not to be taken lightly,” Jackson stated.

“It is imperative that we, as a country, without delay, at least begin the process of exploring legislative safeguards that can punish persons who wilfully utilise such technologies that cause that effect. In other words, you can’t just use the technology, and the information you put out there be damning and you kinda disown it or you take no responsibility and [no] liability follows you. We have to find a way to place on each person some obligation, some liability if they do something that is injurious, that is not correct,” Jackson insisted.

Tufton agreed, saying that “Ultimately, to the extent that people’s reputation are at stake, or important institutions of governance are at stake, there has to be some laws that protect people and those institutions.”

However, he said getting to that point is a function of collaboration and agreement.

“This is one case where all are vulnerable and no one should see it as an advantage to them, because ‘same knife stick sheep stick goat’. It could happen to anybody, and so there is a benefit to collaboration, and we should advocate for that collaboration soonest, so at least from a moral and ethical perspective there is an understanding and an agreement that we need to understand it and its potentials and then, of course, through policy and legal perspectives, it is shared by all, so that all are protected and people compete on merit, as opposed to misinformation, falsehoods, and indeed a sort of toxic, disruptive misleading kind of approach,” Tufton told the Sunday Observer.

Last week, the Media Association of Jamaica (MAJ) cautioned media houses and the general public to “look out for the unscrupulous use of AI for content creation which can be used to mislead and misinform the electorate”.

“Media houses are encouraged to remain vigilant and, if necessary, bolster their editorial and commercial vetting processes to better detect and intercept offending material during this period of heightened activity to keep it out of mainstream media,” the MAJ said in a news release.

“The public is also advised to carefully check the content they consume, along with the sources, and then cross-check with traditionally reputable media brands,” the association advised.

The issue has triggered similar concerns in other jurisdictions, and political watchers have noted that there is growing fear about deepfake’s ability to sway voters, especially given that 2024 is a busy election year.

In England, Sadiq Khan, the mayor of London, and Labour party leader Keir Starmer have been targeted with the use of deepfakes designed to damage them.

Last month, on the eve of the New Hampshire primary in the United States, voters in that state received a robocall purporting to be from US President Joe Biden, even using one of his trademark quips “What a bunch of malarkey!” However, unlike other such calls encouraging people to vote, the call discouraged them from doing so, telling them to save your vote for the November presidential election.

However, America is struggling to deal with the issue and so far only Minnesota, Michigan, and Washington have enacted laws attempting to tackle it. Seven other states have introduced Bills designed to deal with the issue, but those proposals stalled or failed for a number of reasons, among them a recognition that any such regulation would have to survive First Amendment rights challenges.