News
Agent AI Use Guidelines | Google Cloud
Publication date unknown · Discovered google.com
Page published
Publication date unknown · First observed: 15 Oct 2026
Coverage timeline
Single-source advisory — one report is available.
Why it matters
Google Cloud's agentic AI guidelines formalize the threat classes defenders must address — prompt injection, data poisoning, and agent deception — and signal platform-level governance expectations for deploying autonomous agents.
Google Cloud published Agentic AI Responsible Use Guidelines outlining prohibited uses of AI agents, including exploiting vulnerabilities, prompt injection, data poisoning, agent deception, credential misuse, and unauthorized spawning of subagents. The page also warns that agents may act outside instructions and are highly susceptible to external attacks such as prompt injection, recommending human oversight and approval mechanisms.