Meta AI Launches Hebrew Support on WhatsApp with Privacy Concerns Raised

WhatsApp has introduced Meta AI in Hebrew, allowing users to tag the bot in chats and groups, engage in direct conversations, summarize messages, and create content such as stickers. While regular WhatsApp chats remain end-to-end encrypted, interactions with Meta AI differ as the data sent to the bot is accessible to Meta to provide its services. According to Meta's privacy documents, this information may also be used to improve its AI models. Consequently, users are advised against sharing sensitive, personal, business, or commercial information with the bot, especially in work-related groups.

The primary risk for organizations lies in uncontrolled employee use of AI tools, known as Shadow AI, which can lead to the inadvertent sharing of customer data, internal documents, or confidential details with third parties without proper oversight or data processing agreements. Experts recommend establishing clear policies for Meta AI usage on WhatsApp, training employees to avoid inputting sensitive information, and considering blocking AI features in work groups through advanced privacy settings to mitigate these risks.