Skip to main content
Oathera logo Oathera
Use casesFive stories: no identity, a new identity, impersonation, cloning, and a call outside the boundary. PlatformOathera Agent Enrollment, the Oathera Identity Service and Oathera Agent Enforcement. FeaturesWhat Oathera does for you, grouped by purpose. IntegrationsYour identity provider, your MCP servers, your cloud. SecurityHow Oathera fails closed, and what it doesn't do.
Developer sectionComponents, the identity token, standards and settings. Data sheetsThe buyer and developer data sheets, as PDFs. LearnGuides on giving AI agents an identity of their own. GlossaryPrincipal, sponsor, anchor key, key digest, identity token. SimulatorWatch a stolen token, a replayed proof and a tampered call get refused.
About usThe team and the mission behind Oathera. BlogPosts from the Oathera team. CareersOpen roles at Oathera. ContactTalk to the founders, or become a design partner. LinkedIn ↗Oathera on LinkedIn.
Privacy PolicyHow we handle data. Terms of ServiceTerms for using Oathera. Data Processing AgreementOur DPA for customers. Sub-processorsThird parties we rely on.
Become a design partner

Learn · Threats

Confused deputy and prompt-injection privilege escalation in AI agents

A confused deputy is an agent that holds authority a caller does not have and uses it on the caller's behalf. With AI agents, a planted instruction in an email, ticket or web page can steer the agent into using its real access for an attacker. You limit the damage by narrowing what the agent can reach, not by hoping to spot every injected sentence.

On this page

  1. How it happens
  2. Why filtering is not the fix
  3. What limits the damage
  4. How Oathera helps
  5. FAQ

How it happens

An agent reads content written by someone else. A hidden or plain instruction in that content gets treated as a task. The agent then calls a tool with its own permissions. Nothing was hacked in the usual sense: the agent did what its access allowed.

Why filtering is not the fix

The planted text can be worded, encoded or hidden many ways, so a filter is a last layer, never the control. The reliable controls sit outside the model and decide on verified facts: which agent, which person, which operation.

What limits the damage

  • An operational boundary listing only the tools and arguments the agent needs.
  • A check on every call, outside the model, against the agent's verified identity.
  • Delegation to a sub-agent that can only narrow authority.
  • A person's single-use approval for one exact sensitive operation.

How Oathera helps

A steered agent still cannot call an unknown tool, pass arguments outside a tool's allow-list or reach a service outside its boundary, because Oathera Agent Enforcement checks the call and the Oathera Identity Service issues a signed, single-use decision. Model-supplied arguments are labelled as agent content and cannot set protected security fields. That limits what a manipulated agent can do. It does not stop the manipulation, and content is not scanned for injected instructions.

Oathera does not verify people, judge model output, inspect or rewrite prompts, or replace your identity provider. Your identity provider and proofing sources do the first; Oathera is the layer under them for agents.

FAQ

What is a confused deputy in AI agents?

An agent that uses its own authority for a caller who does not have it, often after a planted instruction in content the agent read.

Can you prevent prompt injection?

Not reliably. Plan to limit what an injected agent can do: narrow boundaries and checks on every call, outside the model.

Does Oathera detect prompt injection?

No. Oathera does not inspect prompts or scan content. It limits what any agent, steered or not, can reach.

How can I try these attacks safely?

Play the free Prompt Heist series at oathera.ai/prompt-heist. It runs in your browser.

Related: Enforce least privilege for an AI agent · Operational boundaries, enforced at runtime · Prompt Heist, a free prompt injection game

Become a design partner Watch the simulator More guides

← Back to Learn
Oathera logo Oathera

Every AI agent, sworn in. Oathera gives every AI agent an identity bound to the machine it runs on, with a named person who answers for it, and checked every time it acts.

Product

  • Use cases
  • Platform
  • Features
  • Integrations
  • Security

Developers

  • Developer section
  • Data sheets
  • Learn
  • Glossary
  • Simulator

Company

  • About us
  • Blog
  • Careers
  • Contact
  • GitHub

Legal

  • Privacy Policy
  • Terms of Service
  • Data Processing Agreement
  • Sub-processors
© 2026 Oathera Inc. · All rights reserved
Know Your Agent (KYA)

Cookie preferences

We use cookies to run this site and, with your consent, to understand usage and improve Oathera. Strictly necessary cookies are always on; you can choose whether to allow analytics and marketing cookies below.

Read the privacy policy

  • Strictly necessaryAlways on

    Required for the site to work, security, load balancing, and remembering your cookie choices. These cannot be switched off.

  • Help us measure traffic and see how the site is used, so we can improve it. No personal profiles are built.

  • Used to make messages about Oathera more relevant across other sites. Off unless you turn it on.