The next time a video appears on your phone showing a familiar corporate executive announcing a lucrative giveaway, promising to double your Fuliza limit or inviting you to invest in an exciting new opportunity, take a moment before clicking. The person in the video may not have made the announcement, the offer may not exist, and the website asking for your details may have been created by a fraudster.
As generative artificial intelligence (AI) becomes more accessible, online fraud is taking on a new and more convincing form. Criminals no longer have to rely on poorly written messages, suspicious-looking websites or obvious impersonation attempts. They can now use AI tools to generate realistic videos, imitate voices, manipulate photographs and build convincing copies of legitimate websites to make fraudulent offers appear genuine.
For Kenyan consumers, this presents a growing challenge. Fraudsters are exploiting the trust people have in established brands, particularly those offering financial and mobile money services. Safaricom, its executives, products and promotions can become targets for impersonation, with Fuliza providing an especially attractive subject because of its popularity among customers who rely on the service to meet everyday financial needs.
The danger is not simply that these scams look more professional. It is that they are designed to make people believe they are dealing with a trusted institution at precisely the moment they are being asked to share sensitive information or send money.
When seeing is no longer believing
For years, consumers have been advised to look out for spelling mistakes, strange email addresses, poor-quality graphics and messages promising rewards that seem too good to be true. While these warning signs remain useful, generative AI is making some traditional fraud-detection methods less reliable.
A fraudster can take publicly available footage of a corporate executive and manipulate it to create a video in which the person appears to endorse a promotion or investment scheme. AI-generated voices can make fraudulent phone calls or video announcements sound more authentic. Images, logos, and promotional graphics can be reproduced accurately enough to convince someone scrolling quickly through Facebook, TikTok, WhatsApp, or another social media platform.
The same technology can be used to create cloned websites that closely resemble official corporate pages. A fraudulent page may use familiar branding, professional layouts and language that mirrors genuine customer communications. It may even feature fabricated testimonials, countdown clocks and claims that thousands of Kenyans have already claimed their rewards.
The objective is to remove doubt and create urgency.
A customer who sees what appears to be an announcement from a trusted executive may feel little need to question it. Someone struggling to meet an immediate financial need may be particularly tempted by a message promising to double their Fuliza limit. Others may be drawn to investment opportunities presented as exclusive offers available only to a limited number of customers.
Once trust has been established, the fraudster directs the victim to a fake website, requests personal information, asks for a payment to unlock a reward or attempts to obtain mobile money account credentials and verification codes.
What begins as an apparently harmless click can end in financial loss, identity theft or the compromise of an account.
Why Fuliza impersonation is particularly concerning
Fuliza is a familiar part of Kenya’s mobile financial landscape. Its visibility and relevance to everyday transactions make it an attractive subject for criminals seeking to exploit customers’ financial pressures and expectations.
An advertisement promising an instant increase in a customer’s Fuliza limit can appear especially appealing. A fraudulent post may claim that customers can double their limits by registering through a special link, paying a small processing fee or providing their M-PESA details for verification.
Other schemes may promise cash rewards, free airtime, smartphones or investment returns supposedly backed by Safaricom. Some may combine several of these claims, using a giveaway to attract attention before directing users towards a fraudulent financial offer.
Consumers should understand that a video featuring a Safaricom executive, a familiar logo or an apparent customer testimonial is not proof that an offer is genuine. Nor does a high number of views, likes or comments establish legitimacy. Engagement can be manipulated, and comments praising a promotion can be fabricated or generated through coordinated accounts.
Any unexpected offer promising to double a Fuliza limit, guarantee an investment return or provide a reward in exchange for sensitive information should be treated with caution and independently verified through Safaricom’s official communication channels.
Customers should never assume that an offer is legitimate simply because it appears in a sponsored social media advertisement. Paid promotion does not, by itself, guarantee that the advertiser is who they claim to be.
The pressure to act quickly is part of the trap
One of the most effective weapons used by digital fraudsters is urgency. Fake promotions frequently claim that an offer expires within minutes, that only a few customers qualify or that immediate registration is necessary to avoid missing out.
These messages are designed to discourage verification. When people believe an opportunity is about to disappear, they are more likely to act on instinct rather than examine the details.
The best defence is to slow down.
Before acting on a promotion, customers should independently visit Safaricom’s official website or verified social media accounts to establish whether the offer has been announced. They should avoid relying on links shared by strangers, forwarded WhatsApp messages or accounts that merely use Safaricom’s name and branding.
Where an offer involves Fuliza, customers should use established, official M-PESA and Safaricom channels to check the relevant information. If an announcement cannot be independently confirmed, it is safer to ignore it.
Customers should also avoid entering their M-PESA PINs, passwords, one-time verification codes or other confidential information on websites reached through unsolicited promotional links. Such details should never be disclosed to someone claiming they are needed to release a giveaway or increase a loan limit.
A legitimate-looking website address deserves particular scrutiny. Fraudsters often use addresses that differ only slightly from genuine ones, adding a word, changing a character or using an unfamiliar domain. The presence of a padlock symbol in a browser is not sufficient proof that a website belongs to Safaricom; it generally indicates that the connection is encrypted, not that the organisation operating the site is legitimate.
Consumers, platforms and companies all have a role to play
The growing sophistication of AI-enabled fraud means that consumer awareness must evolve alongside the technology.
Individuals should make verification a routine habit rather than something they do only when a message looks suspicious. They should report suspected fraudulent advertisements and impersonation accounts to the relevant social media platforms and notify Safaricom through its official customer support channels. Anyone who has shared sensitive information or lost money should act promptly to secure the affected account, contact the relevant service provider and report the incident to the appropriate authorities.
Social media companies also have a responsibility to respond to the misuse of their advertising and content distribution systems. Their detection mechanisms should be strengthened to identify impersonation, suspicious landing pages and coordinated fraudulent campaigns. Reporting mechanisms must be accessible, and complaints involving financial scams should receive timely attention.
Businesses, meanwhile, must recognise that protecting customers extends beyond securing their own systems. Their names, executives, products and reputations can be misused in environments they do not directly control. This calls for clear public communication, rapid responses to emerging scams and regular reminders explaining how customers can distinguish genuine offers from fraudulent ones.
Public education is particularly important because the consequences extend beyond individual financial losses. Successful impersonation campaigns can undermine confidence in legitimate digital financial services, damage corporate reputations and discourage consumers from using technologies that offer real economic benefits.
Kenya has made significant progress in mobile financial services and digital commerce. Preserving that progress requires consumers to be informed, companies to remain vigilant and digital platforms to take the misuse of their services seriously.
AI is a tool, but trust remains the target
Generative AI is not inherently fraudulent. It can support innovation, improve customer service and make digital communication more efficient. The problem arises when criminals use its capabilities to deceive people at scale.
The technology may change, and the scams may become more sophisticated, but the underlying tactic remains familiar: create a believable story, exploit trust and persuade the victim to act before thinking.
The practical lesson for Kenyans is straightforward. Do not trust an online offer solely because it features a recognisable face, sounds like a familiar voice or carries the branding of a respected company. Verify the source independently, be sceptical of extraordinary promises and never surrender confidential information under pressure.
A convincing deepfake can imitate an executive. A cloned website can imitate a company. A fraudulent advertisement can imitate a genuine promotion. But none of these things makes the offer real.
In an increasingly AI-driven digital environment, the most valuable security habit may be the simplest one: pause, verify and only then act. That brief moment of caution could be the difference between enjoying a legitimate service and becoming the next victim of digital fraud.