Palo Alto Networks researchers show how attackers could exploit AI agents on Google's Vertex AI to steal data and break into restricted cloud infrastructure.

Read original →