Meta Deletes 6.8 Million WhatsApp Accounts Tied to AI Scams, Boosts User Protections

Meta has carried out its biggest cybersecurity enforcement on WhatsApp, deleting 6.8 million accounts associated with AI-driven scam operations in the first half of 2025. The action focuses on criminal networks exploiting AI for fraudulent schemes. In collaboration with OpenAI, Meta aims to prevent AI-generated messages from reaching potential victims. Investigations revealed that many of the removed accounts were linked to organized crime hubs in Southeast Asia, where authorities reported operations under conditions akin to modern slavery.

Figure 1. WhatsApp: Preferred Platform for Cybercriminals

Next-Generation Detection Systems Fight Cyber Threats

WhatsApp now uses machine learning to detect suspicious behavior patterns before criminals can activate accounts. Enhanced security features include alerts when unknown contacts try to add users to group chats—a common tactic for expanding criminal reach. Meta emphasizes that proactively identifying and removing accounts before scam networks can operate marks a significant move toward preventive, rather than reactive, cybersecurity. Figure 1 shows WhatsApp: Preferred Platform for Cybercriminals.

Uncovering Criminal Networks

The investigation revealed a Cambodian criminal group using ChatGPT to create initial contact messages for victims. These AI-generated texts, containing WhatsApp links, direct targets to Telegram, where they are tasked with activities like interacting with TikTok videos in exchange for promised rewards [1]. The criminals use sophisticated social engineering, including fake success screenshots, to persuade victims to deposit cryptocurrency for access to supposedly higher-paying tasks.

Key Challenges in Cybersecurity

The enforcement action underscores the growing scale of cybercriminal networks using AI to expand their operations. While companies like Meta employ AI for fraud detection, these criminal groups exploit the same technology to craft convincing messages and automate targeting. Countries such as Myanmar, Cambodia, and Thailand have become key hubs for these activities, with local authorities struggling to contain networks that grew during the pandemic and have defrauded victims of tens of billions of dollars globally. Consumer rights organization Which? recognizes Meta’s efforts but urges broader cybersecurity measures across all its platforms, stating, “Meta must do much more to stop these criminals across all its platforms.”

The cybersecurity challenge also highlights gaps in regulatory frameworks, as UK regulator Ofcom faces calls to tighten enforcement of the Online Safety Act amid widespread cyber fraud. Although removing 6.8 million accounts shows WhatsApp’s dedication to security, the rapidly evolving landscape of AI-driven threats indicates that this action is only the start of an ongoing fight against digital fraud.

References:

1. https://cybermagazine.com/news/netscout-ai-supercharges-ddos-threats-globally

Cite this article:

Janani R (2025), WhatsApp AI Scams: Meta Deletes 6.8 Million Accounts, AnaTechMaz, pp. 311