WhatsApp Vulnerability: AI Agents Can Spam Contacts

Aug 6, 2026

72 min

Aug 6, 2026 10:31

WhatsApp warning: AI agents can spam all your contacts — here’s how

## AI Exploitation in Browsers

Recent findings from cybersecurity firm Zenity reveal that AI agents embedded in web browsers can be manipulated to perform harmful actions, such as spamming all contacts in a user's WhatsApp account. This discovery was presented at the Black Hat conference in Las Vegas.

## How the Attack Works

The attack targets OpenAI's Atlas browser, which includes an AI agent capable of executing tasks online. Researchers demonstrated that by embedding hidden instructions in a webpage, the AI could be tricked into sending spam messages to all WhatsApp contacts of a user. These instructions bypassed Atlas's safety filters by using non-English languages and combining them with legitimate commands.

## Broader Security Concerns

Zenity's study identified over 20 security vulnerabilities across AI-enabled browsers and extensions from major tech companies. These issues could lead to unauthorized access to personal data and accounts. Despite having robust protections, Atlas was still compromised, highlighting the need for improved security measures.

## Challenges in Addressing the Issue

The structural design of AI agents, which allows them to operate across different websites, makes them susceptible to prompt injection attacks. This vulnerability undermines traditional browser security protocols.

## OpenAI's Response and Future Plans

OpenAI has acknowledged the issue and is working on improving protections against prompt injection. The company plans to discontinue the Atlas browser on August 9, shifting focus to a ChatGPT extension for Chrome and enhanced browsing features in its desktop app.

Read the full story at the source

What you need to know to get Emirates ID?

Leave your details and get a guide as a gift to avoid mistakes

Article contents