Deepfakes, AI fuel new wave of cyber attacks on businesses

Raphael John-Lall

Busi­ness own­ers are be­ing urged to strength­en their cy­ber­se­cu­ri­ty de­fences as in­creas­ing­ly so­phis­ti­cat­ed cy­ber­at­tacks—dri­ven by ar­ti­fi­cial in­tel­li­gence and deep­fake tech­nol­o­gy—pose a grow­ing threat to op­er­a­tions and fi­nan­cial sta­bil­i­ty.

Man­ag­ing Di­rec­tor of Pina­ka Con­sult­ing, Shi­va Bisses­sar, warned that cy­ber­crim­i­nals are rapid­ly evolv­ing their tac­tics, us­ing more con­vinc­ing and tech­no­log­i­cal­ly ad­vanced meth­ods to ex­ploit vul­ner­a­bil­i­ties with­in or­gan­i­sa­tions.

Speak­ing to the Busi­ness Guardian, Bisses­sar high­light­ed deep­fakes as one of the most press­ing risks fac­ing busi­ness­es to­day. He al­so ref­er­enced his pre­sen­ta­tion at the CAN­TO Con­nect 2026 and its 42nd AGM, held from Feb­ru­ary 1 to 3 at the Hy­att Re­gency Ho­tel in Port of Spain, where the is­sue was a cen­tral fo­cus.

CAN­TO Caribbean, in a so­cial me­dia post on the event, not­ed that Bisses­sar ad­dressed the grow­ing threat posed by deep­fakes—AI-gen­er­at­ed con­tent ca­pa­ble of ma­nip­u­lat­ing au­dio, video and im­ages in ways that un­der­mine trust, se­cu­ri­ty and de­ci­sion-mak­ing across gov­ern­ments, busi­ness­es and so­ci­ety.

That ses­sion ex­am­ined what deep­fakes are, why they mat­ter, and how busi­ness own­ers, or­gan­i­sa­tions and pol­i­cy­mak­ers can re­spond in an in­creas­ing­ly com­plex dig­i­tal en­vi­ron­ment.

“So, what is the at­tack evo­lu­tion that we’re see­ing right now? Deep­fakes, AI-gen­er­at­ed con­tent, it’s a threat that is out there right now, and we’re see­ing it mov­ing. You guys would have seen the videos with Will Smith two years ago eat­ing spaghet­ti, where he looks re­al­ly con­tort­ed and his face is all over the place, to now where we have re­al­is­tic deep­fakes.”

He ex­plained that cy­ber­crim­i­nals are shift­ing away from tra­di­tion­al so­cial en­gi­neer­ing tech­niques to more ad­vanced, AI-en­hanced de­cep­tion strate­gies.

“And the at­tack evo­lu­tion is go­ing to be now, mov­ing so­cial en­gi­neer­ing from your tra­di­tion­al meth­ods of try­ing to fool peo­ple, to try­ing to fool peo­ple with more con­vinc­ing deep­fakes. So so­cial en­gi­neer­ing is not a tech­ni­cal term, it’s some­thing that you do all the time. You come and you meet an au­di­ence, you so­cial en­gi­neer your way in­to a room. So­cial en­gi­neer­ing is how we con­vince peo­ple to trust us, to car­ry out some ac­tion that we want them to car­ry out. It’s not tech­ni­cal.”

Bisses­sar warned that this evo­lu­tion is al­ready af­fect­ing busi­ness­es through schemes such as busi­ness email com­pro­mise, where fraud­sters im­per­son­ate ven­dors or ex­ec­u­tives to ma­nip­u­late em­ploy­ees in­to au­tho­ris­ing pay­ments or grant­i­ng ac­cess.

“We have now the ad­vent of at­tacks that are based on busi­ness email com­pro­mise. So, a busi­ness email com­pro­mise is where you would see some­one pre­tend­ing to be a ven­dor. They will send you an email, and they are try­ing to con­vince you to click on that email and add them to the ven­dor list, or send a wire to an ex-dif­fer­ent ac­count that you’re nor­mal­ly ac­cus­tomed to. That busi­ness email com­pro­mise is very preva­lent in T&T.”

He stressed that es­tab­lished in­ter­nal con­trols—par­tic­u­lar­ly those gov­ern­ing fi­nan­cial trans­ac­tions—must not be by­passed, even when com­mu­ni­ca­tions ap­pear le­git­i­mate.

“You have to go through a process. But this is the game. These guys are so good at try­ing to get you to by­pass your ex­ist­ing work­flows. This is what they try to do. So, they would send you an email, pre­tend to be a sup­pli­er, hey, click on this link to add them to the list. Click on this link to send mon­ey to a dif­fer­ent ac­count.”

New and Emerg­ing Threats

Bisses­sar al­so point­ed to a grow­ing range of AI-en­abled threats, not­ing that cy­ber­crim­i­nals are re­fin­ing their tac­tics us­ing more so­phis­ti­cat­ed scripts and de­cep­tion tech­niques.

“What are some of these AI-en­abled tech­niques? We’ve seen the tech­ni­cal mea­sures that bad ac­tors are try­ing to per­pe­trate against the good ac­tors, where are the good ac­tors, or their po­ten­tial vic­tims. The ar­eas that they would ex­ploit would be that pre­text script that they have to try to con­vince some­body that they are some­one else. Those scripts will now im­prove with AI.”

He high­light­ed “fi­nan­cial sex­tor­tion” as an­oth­er emerg­ing con­cern, par­tic­u­lar­ly af­fect­ing younger users.

“This is where per­sons try to prey on teenagers and get them to sub­mit some com­pro­mis­ing pho­to of them­selves, and they will use that now as lever­age to ask them for mon­ey once they’ve fall­en vic­tim to that scheme. These at­tacks are not per­pe­trat­ed by in­di­vid­u­als. There are groups known for these types of at­tacks, and they are based in cer­tain re­gions of the world.”

He not­ed sim­i­lar­i­ties with ro­mance scams, which have al­ready been flagged lo­cal­ly by the Fi­nan­cial In­tel­li­gence Unit of Trinidad and To­ba­go.

“Ro­mance scams have been recog­nised by the Fi­nan­cial In­tel­li­gence Unit of Trinidad and To­ba­go (FI­UTT) as some­thing tak­ing place in Trinidad and To­ba­go of sig­nif­i­cance. They have da­ta to sup­port that. They’ve is­sued a re­port. They’ve is­sued warn­ings on that, where­as the fi­nan­cial sex­tor­tion is some­thing that ap­plies to the youth. But this type of ex­tor­tion, if we look at it from that gen­er­al word of ex­tor­tion, it is some­thing that takes place glob­al­ly, and it has been tak­ing place in T&T.”

Bisses­sar added that deep­fake-en­abled fraud is no longer the­o­ret­i­cal, cit­ing a high-pro­file case in Hong Kong where a com­pa­ny em­ploy­ee was tricked dur­ing a video call in­to trans­fer­ring mil­lions of dol­lars.

“Deep­fake at­tacks did take place in Hong Kong maybe a cou­ple years ago, where the con­troller of ac­counts was on a con­fer­ence call that he thought was his ex­ec­u­tives, and they were ba­si­cal­ly scammed out of…he sent mon­ey. He fol­lowed the in­struc­tions re­ceived on that video call, and he sent mon­ey to the tune of, I be­lieve, $25 mil­lion. So, this is not fic­tion.”

He al­so warned that ran­somware at­tacks are be­com­ing more or­gan­ised, with cy­ber­crim­i­nals op­er­at­ing in struc­tured net­works re­sem­bling le­git­i­mate busi­ness ecosys­tems.

“Ran­somware is be­ing per­pe­trat­ed by a group of bad ac­tors. It’s an ecosys­tem. It’s a busi­ness ecosys­tem. Just as you have economies of scale in the every­day world, these guys have classed them­selves in­to ar­eas of spe­cial­i­sa­tion.”

Call for Col­lab­o­ra­tion

Against this back­drop, Bisses­sar called for stronger col­lab­o­ra­tion be­tween the pub­lic and pri­vate sec­tors to counter the grow­ing cy­ber threat.

“The bad ac­tors are cre­at­ing ecosys­tems. What are the good ac­tors do­ing? Do we tru­ly have pub­lic-pri­vate part­ner­ship in Trinidad and To­ba­go when it comes to cy­ber­se­cu­ri­ty? We’ve heard some great plans about what the Telecom­mu­ni­ca­tions Au­thor­i­ty of Trinidad and To­ba­go (TATT) is do­ing. But I would en­cour­age you to part­ner with the pri­vate sec­tor as well. Part­ner with some of the small­er pri­vate sec­tor en­ti­ties as well, be­cause we have cer­tain ex­per­tise that we could lend to this type of sit­u­a­tion.”

As cy­ber threats be­come more so­phis­ti­cat­ed, he stressed that vig­i­lance, ro­bust in­ter­nal con­trols and col­lab­o­ra­tion will be crit­i­cal in safe­guard­ing busi­ness­es from in­creas­ing­ly com­plex dig­i­tal risks.