Meta's Hatch AI Agent Exposes Sensitive Data During Internal Testing
Meta announced Hatch, an AI agent that performs autonomous tasks like sending emails, selling vehicles, and booking travel. The company faced internal concerns that the technology mishandles access to sensitive personal data during testing, raising privacy and security questions.
What happened
Meta announced an AI agent called Hatch that can autonomously send emails, sell vehicles, and book travel on behalf of users.
Internal concerns were raised that the technology mishandles access to sensitive personal data during testing.
Reported impact
Users are required to grant the agent access to personal information and banking credentials to perform these autonomous tasks.
Index analysis
Analysts raised questions about privacy, consumer trust, and the computational resources required for routine daily activities.
The deployment timeline and availability in the European Union and Greece remain unclear due to strict GDPR and AI Act regulations.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.