In an era where artificial intelligence has become both a blessing and a battleground, Google is doubling down on its AI arsenal to combat the dark side of digital innovation—fraud, scams, and phishing operations that plague millions of users daily.
With Gemini, its next-generation AI system, Google is fortifying its ecosystem—from Chrome to Android—against an evolving wave of cyber deception. And while AI has fueled the sophistication of online scams, it may also be the key to neutralizing them.
Fraudulent websites, cloned customer service pages, and impersonation scams are nothing new. What is new is the scale and speed with which bad actors use AI to launch attacks, often outpacing traditional security systems.
“AI has become the cybercriminal’s weapon of choice,” says Dr. Nader El-Kady, a digital security consultant for Egypt’s Ministry of Communications and Information Technology. “What used to take hours to code and execute now takes seconds with generative AI. Google’s pivot to using Gemini to fight back is a timely and necessary evolution.”
The Gemini Nano model, a lighter, on-device AI engine, is now embedded into Chrome and Android. It operates locally on users’ phones—without sending data to Google’s servers—flagging phishing attempts, dangerous links, and fraudulent customer service pages in real time. This preserves user privacy while enhancing proactive threat detection.
According to Google’s internal analysis, AI-assisted efforts have led to a 20-fold increase in scam detection, resulting in an 80% drop in fake and malicious pages appearing in search results.
A spokesperson from Google MENA, in exclusive comments to MEO, noted that this update has already been rolled out in English, Arabic, and French across Chrome in Egypt, Saudi Arabia, and the UAE. “Our AI is trained on regional threat vectors. From SMS loan scams in Cairo to phishing banking emails in Dubai, Gemini is now context-aware across multiple dialects and fraud patterns.”
Gemini’s rollout goes beyond browsing. Android users will soon experience AI-powered fraud detection across apps like Google Messages and the default Phone app. These features aim to replicate and improve upon the role of third-party tools such as Truecaller, which dominate caller ID and spam filtering in markets like Egypt and India.
In high-risk regions, Android will now automatically flag suspicious numbers, highlight scam calls, and label phishing SMS in real time. Experts predict this could significantly reduce cybercrime incidents, especially in emerging markets where digital literacy remains uneven.
The Middle East is increasingly a target for AI-enhanced fraud, particularly in nations undergoing digital transformation. In Egypt alone, the National Telecom Regulatory Authority (NTRA) recorded a 45% year-on-year increase in consumer fraud complaints in 2024, much of it attributed to impersonation scams using AI-generated content.
Cybersecurity expert Mariam Khaled, Director at Cairo-based infosec firm CyGuard, believes the solution is twofold: “AI will be critical, but user education must follow. If people understand the red flags and the tools at their disposal—like Gemini—they’ll be far less vulnerable.”
Despite its promise, experts caution that Gemini is not a silver bullet. As AI arms both defenders and attackers, the cat-and-mouse game intensifies.
“AI will never eliminate fraud completely,” said Fadi Dabbous, Senior Policy Analyst at the Arab Digital Economy Forum.“But it can significantly shift the balance of power—especially if embedded into platforms used by billions, like Chrome and Android.”
Google’s anti-fraud push aligns with its broader AI footprint in the MENA region. Recent launches include:
- Arabic Gemini AI expansion into Google Search and Assistant.
- Cyber hygiene campaigns in partnership with Egypt’s Ministry of Education and digital startups.
- Ongoing development of localized AI datasets to better detect Arabic-language phishing and scam variants.
Sources at Google confirmed to MEO that a dedicated AI fraud detection center is under discussion for launch in Dubai Internet City, which would focus on regional threat mapping and Arabic-language scam detection.
