Technology ❯ Computer Science ❯ Machine Learning ❯ Large Language Models
The company is prioritizing defenses against indirect prompt injection that can hijack web agents.
