Egypt is the second most ransomware-targeted country in Africa, accounting for 13 percent of all cyberattacks on the continent. As of 2025, the nation loses an estimated USD 4 billion (EGP 207.52 billion) to cybercrime every year, according to Interpol’s Africa Cyberthreat Assessment Report 2025.
Such statistics might have seemed troubling before artificial intelligence (AI) entered the picture. Currently, with AI tools, scams are faster, cheaper, and nearly impossible to detect. The threat has entered a new phase, and most internet users have no idea what they are up against.
AI-enabled fraud surged 1,210 percent in 2025, far outpacing the 195 percent growth in traditional fraud over the same period. The era of the obvious scams with misspelled emails or suspicious wire transfer requests is over. AI has quietly erased the telltale signs that once protected users, and the consequences are staggering.
According to projections from the Deloitte Center for Financial Services, published in May 2024, generative AI-driven fraud losses in the U.S. alone are expected to grow from USD 12.3 billion (EGP 638.12 billion) in 2023 to USD 40 billion (EGP 2 trillion) by 2027.
Cybercrime’s global rise hits close to home
The Federal Bureau of Investigation (FBI)’s Internet Crime Complaint Center recorded more than one million complaints in 2025, with reported losses totaling USD 20.8 billion (EGP 1.078 billion) in the U.S., which is the highest ever recorded, and a 26 percent increase from 2024. Nearly 85 percent of that damage came from cyber-enabled fraud.
The problem is not confined to countries with wealthy economies. Interpol’s 2025 Africa Cyberthreat Assessment found that scam notifications rose by up to 3,000 percent in some African countries, with Egypt sitting at the center of that surge.
Interpol specifically flagged Egypt as one of Africa’s most heavily targeted nations, with social engineering, phishing via messaging platform WhatsApp, and fake investment schemes among the most widespread tactics.
In 2024 alone, Positive Technologies, an information security company, identified over a hundred dark web listings offering stolen Egyptian databases for sale.
New tools of deception
Understanding the scale of the problem requires understanding the tools. While AI did not introduce new types of scams, it has, however, made old ones easier to commit, virtually undetectable, and infinitely scalable.
Generative AI can now produce convincing deepfake videos and audio recordings of public figures or copy the appearance, voice, and mannerisms of virtually anyone using existing video or audio footage. Financial losses from deepfake-enabled fraud exceeded USD 200 million (EGP 10.4 billion) in Q1 2025 worldwide.
In one of the most studied fraud cases of the decade, a finance worker at global engineering firm Arup was tricked into wiring USD 25.5 million (EGP 1.3 billion) after participating in a video call featuring AI-generated likenesses of senior executives.
Via business email compromise (BEC), criminals can use AI to impersonate executives, generate fake invoices, and issue fraudulent payment instructions. BEC costs businesses at least USD 3 billion (EGP 155.6 billion) globally in 2025. AI voice cloning has supercharged these attacks by simulating voices that are nearly impossible to distinguish from a real one.
Replicating a person’s voice can take as little as three seconds of audio using voice cloning. One in ten adults globally has experienced an AI voice scam, and 77 percent of those targeted reported losing money as of September 2025. The most common scam is receiving a call from a loved one who has an emergency and asks for an urgent money transfer.
Another infamous method is investment or romance-based scamming, where scammers use AI to build trust over weeks through fake personas, then direct victims into fraudulent investment platforms. AI-generated fake celebrity endorsement videos drove investment scam losses exceeding USD 632 million (EGP 32.7 billion) in the U.S. alone in 2025. In March 2026, Interpol flagged fake investment schemes as one of the most common and damaging scam categories worldwide.
Another type of scam is a fraudulent email, which used to be identifiable through classic indicators, such as poor grammar and generic greetings, but AI has changed that. Over 82 percent of phishing emails are now created with the help of AI, allowing fraudsters to craft convincing, personalized messages faster. Phishing reports increased 466 percent between Q1 2024 and Q1 2025, according to a report by Gen Threat, a cybersecurity company. In Egypt, phishing via WhatsApp is among the most widespread tactics.
AI-powered bots can also conduct thousands of fraudulent transactions simultaneously, harvesting credentials and processing fake refund requests at a scale no human operation could match. Among major retailers, one study recorded a 330 percent increase in automated bot fraud in two months.
A dangerous confidence gap
According to Sift, a fraud prevention company, seventy percent of online users said that it had become more difficult to identify scams in 2025 compared to the prior year. Yet, one-third still say they are confident they could detect an AI-generated scam, presenting a dangerous disconnect.
Currently, 78 percent of users worldwide open AI-generated phishing emails, and 21 percent click on malicious links, according to another report by Sift. Meanwhile, in Africa, Interpol has noted that 75 percent of African countries report the need for improved legal frameworks for prosecuting cybercrime.
While the same AI being weaponized by criminals is being deployed in defense by banks, cybersecurity firms, and law enforcement, AI-generated fraud remains on the rise.
Comments (0)