Imagine receiving a message from someone you trust.
The voice sounds familiar, the story is convincing, and the request for money seems urgent.
You send the cash, believing you have helped a relative, supported a promising investment or responded to a legitimate business opportunity.
Only later do you discover that the person you thought you were dealing with may never have contacted you at all.
Across Africa, the tools used to manufacture such deception are becoming more sophisticated.
Artificial intelligence can now help criminals imitate voices, generate convincing images, produce polished messages and automate conversations, making it increasingly difficult for ordinary people to distinguish between a genuine opportunity and a carefully constructed fraud.
The scale of the threat is captured in a warning by the International Criminal Police Organization, INTERPOL, which says artificial intelligence was linked to 55 per cent of reported cybercrimes across Africa in 2025.
Its African Cyberthreat Assessment Report 2026, released in August, drew on survey data from 36 African member countries and described a criminal landscape increasingly driven by technology, organised networks and cross-border operations.
The financial toll is equally concerning.
INTERPOL reported that cybercrime-related losses across the continent had more than doubled, rising from 192 million US dollars in 2024 to 484 million dollars in 2025.
Online scams remained the most frequently reported form of cybercrime, with fraudsters exploiting social media, mobile money platforms and digital financial services to reach potential victims.
But behind the statistics are people who have lost savings, businesses that have been deceived and families left trying to recover money they may never see again.
The problem is not confined to sophisticated hackers operating from hidden locations.
Some schemes begin with something as ordinary as a social media advertisement, a message in a WhatsApp group or a seemingly harmless invitation to join an investment opportunity.
In Kenya, for example, 27 people were arrested in February 2026 during an international operation targeting online investment scams, mobile money fraud and fraudulent mobile loan applications.
The arrests formed part of Operation Red Card 2.0, which involved law enforcement agencies from 16 African countries.
According to INTERPOL, the Kenyan schemes relied on messaging applications, social media and fabricated testimonials to persuade people to invest in purported opportunities associated with well-known global corporations.
Victims were encouraged to begin with relatively small amounts, sometimes as little as 50 US dollars, on the promise of attractive returns.
Once they had deposited their money, they were shown fabricated account statements or online dashboards suggesting their investments were growing.
When they attempted to withdraw their funds, their requests were blocked.
The scheme illustrates how online fraud can exploit a basic human instinct: the desire to trust what appears familiar and promising.
A recognisable company name, professional-looking website or convincing testimonial can make an invented opportunity seem legitimate.
A small initial deposit can also lower a person's guard, creating the impression that the risk is manageable before larger sums are requested.
The February operation extended beyond Kenya.
Authorities across the participating countries made 651 arrests, identified 1,247 victims and uncovered fraud schemes linked to more than 45 million US dollars in losses.
They also recovered more than 4.3 million dollars and seized thousands of electronic devices.
The operation targeted several forms of cybercrime, however, and INTERPOL did not attribute every scheme uncovered to artificial intelligence.
That distinction matters because AI is not responsible for every online scam.
Many criminals continue to rely on familiar techniques such as phishing, identity theft, fake websites and social engineering.
What AI changes is the speed, scale and sophistication with which some of these methods can be deployed.
Consider how easily a fraudster can now create a convincing message.
In the past, a poorly written email riddled with grammatical mistakes might have alerted a recipient to a scam.
Generative AI can produce fluent, professional messages in different languages, making fraudulent communications harder to distinguish from genuine ones.
Voice-cloning technology adds another layer of deception. Using recordings of a person's speech, certain tools can generate audio that resembles their voice.
Deepfake technology can manipulate or create videos that appear to show someone saying or doing something they never did.
Criminals can use such material to impersonate relatives, business executives, public figures or supposed investment experts.
A fraudster might, for example, send an employee an audio message that appears to come from a manager requesting an urgent payment.
Another might use a fabricated video of a familiar public figure to promote a fraudulent investment scheme.
Someone else could create a false online identity, build a relationship with a victim over several weeks and eventually request money.
These examples illustrate possible methods rather than a claim that each is behind a particular African arrest.
The central risk is that AI can make deception more convincing, while social media gives criminals access to large audiences and digital financial services provide channels through which money can be moved.
INTERPOL's wider findings suggest that these threats are becoming increasingly organised.
In October 2026, the organisation announced the results of Operation Serengeti III, a separate crackdown involving 21 countries.
The operation resulted in 1,462 arrests, the identification of 65,929 victims and the recovery of more than 105 million US dollars in illicit funds.
Investigators linked the criminal activity to losses exceeding 385 million dollars.
The operation targeted online scams, business email compromise, ransomware and digital extortion, among other cyber-enabled offences.
The figures offer a glimpse of the scale of enforcement required to disrupt networks that can operate across several countries at once.
A scammer may contact a victim in one country, use a website hosted elsewhere and route payments through accounts or intermediaries in another jurisdiction.
Tracing the money and identifying those responsible can require cooperation between police, financial institutions, telecommunications companies and technology platforms.
For ordinary users, however, the first line of defence often begins with a simple pause before acting.
If a relative sends a voice note asking for emergency financial assistance, call them using a number you already know rather than relying on the contact details in the message.
If a well-known businessperson appears to endorse an investment, check the person's verified channels and the company's official website.
A familiar face or voice is not proof that an offer is genuine.
Investment opportunities also deserve careful scrutiny.
Promises of guaranteed profits, unusually high returns or pressure to deposit money immediately should raise suspicion.
Verify whether the company is authorised by the relevant financial regulator, and do not rely on testimonials, screenshots or online dashboards as proof that an investment exists.
The same caution applies to mobile money and digital banking.
Never disclose a PIN, banking password or one-time verification code to someone contacting you unexpectedly.
Avoid clicking suspicious links or installing applications from unknown sources, particularly when the person recommending them is promising easy money or demanding urgent action.
If you suspect that you have been scammed, preserve the messages, telephone numbers, screenshots and transaction references.
Contact your bank or mobile money provider immediately if money has been transferred or your account may have been compromised.
Report the incident to the relevant police or cybercrime reporting authority.
In Kenya, suspected cyber incidents can be reported to the National Kenya Computer Incident Response Team Coordination Centre, while suspected criminal fraud can also be reported to the Directorate of Criminal Investigations.
Yet asking individuals to be more vigilant is only part of the solution.
Social media platforms must respond effectively to fraudulent advertisements and impersonation accounts.
Banks and mobile money providers need systems that can identify suspicious transactions, protect compromised accounts and give victims clear channels for reporting fraud.
Telecommunications companies can help detect suspicious SIM activity, while law enforcement agencies need the skills and resources to investigate digital evidence and cooperate across borders.
Cybersecurity specialists also face the challenge of keeping pace with technology that evolves quickly.
Detecting an AI-generated voice or video is not always straightforward, and visual clues alone may be unreliable. Verification systems, secure account access and rapid reporting mechanisms are therefore important alongside public awareness.
INTERPOL has called for stronger cross-border cooperation, improved digital-forensic capabilities and greater investment in law-enforcement readiness to address the evolving threat.
Its findings suggest that as digital services expand, the ability to investigate and disrupt cybercrime must develop alongside them.
Artificial intelligence itself is not the enemy.
The same technology can support healthcare, education, research, business and communication.
The danger emerges when criminals use it to manufacture identities, manipulate trust and make fraudulent offers appear genuine.
For Africa's growing digital economy, the challenge is to protect the opportunities technology creates without allowing deception to flourish unchecked.
That requires action from financial institutions, technology companies, regulators and law enforcement, as well as informed users who know when to question what appears on their screens.
