Enzai MCP server
Connects to your AI.
Any MCP-compatible agent gets the same governed connection, right where your teams already work.
Some of our integrations
Workflow tools
Jira, ServiceNow
Model providers
Anthropic, OpenAI
Agent tools
Copilot Studio, Claude Enterprise
Development environments
Azure AI Foundry, Amazon Bedrock, Databricks

Glossary

Agent Goal Hijack (ASI01)

Agent Goal Hijack (ASI01)

Agent Goal Hijack (ASI01)

A vulnerability where an attacker manipulates an agent's objectives or decision pathways to redirect its autonomous behavior toward unintended outcomes.

A vulnerability where an attacker manipulates an agent's objectives or decision pathways to redirect its autonomous behavior toward unintended outcomes.

Updated

ASI01 is the agentic evolution of prompt injection. Because agents cannot reliably distinguish between a developer's instructions and untrusted data, an attacker can hijack the goal state. This results in the agent pursuing a new, malicious mission (e.g., exfiltrating data) while believing it is still following its original sanctioned plan.

Example

A travel agent reads a poisoned flight review that contains a hidden instruction to send all booking confirmations to attacker@email.com, which the agent then does autonomously.

Govern the AI you actually use.

Govern the AI you actually use.

Govern the AI you actually use.

Start with what’s in use.

Start with what’s in use.

Start with what’s in use.

Book a demo and we’ll show you how Enzai finds the AI in use, gives each system an owner and keeps the record.

Book a demo and we’ll show you how Enzai finds the AI in use, gives each system an owner and keeps the record.