Skip to content
← Writing
Writing · October 2026 · 2 min read

Your AI agent has more access than you think

Give an AI agent a broad role for one task and it can use that role everywhere. Grant only what the job needs, and check what came bundled.

Abhijit Chatterjee Chattopadhyay

The Logic Atelier logo above the words: Your AI agent has more access than you think.

Your AI agent was set up to do one job. It can probably do far more.

The role nobody meant to grant

Agents get access the quick way. Someone needs it working today, so the agent gets an existing role, an admin key or a service account that already has what it needs, and a lot more.

AWS describes the trap in the same Trust in Depth (opens in a new tab) guidance I quoted on Monday: "Grant an admin role for one task, and the agent can inherit permissions across your whole environment." It also warns about "capabilities no one granted on purpose, the ones bundled with a tool or a role."

The agent does not need to be malicious for this to hurt. A wrong step, a confusing instruction or a prompt injection is enough, if the access is there.

Why it matters more for agents

A person with too much access still has to decide to use it. As AWS notes in the same piece, an agent finds its own path to finish a task, picking tools and steps no developer scripted. Whatever it can reach is on the table. And unlike a person, an agent often has no identity of its own, so nobody can tell afterwards which one acted.

So scope the agent before it runs. AWS calls scoping "where to invest the most effort. Problems caught here are cheap to fix."

Three checks before an agent goes live

  1. What is its job, in one sentence? "Answer delivery questions for existing orders." Not "help customers".
  2. What does that job need? List the data, tools and actions. Read orders, yes. Issue refunds, no.
  3. What did it get anyway? Compare that list with the access it actually holds. Everything extra comes off, including what arrived bundled with a tool or a role.

The gap between check 2 and check 3 is your real risk.

Do this today

  • Pick your most-used agent and write its job in one sentence.
  • Pull the list of permissions its account or role actually has.
  • Remove anything the sentence does not need.
  • Give each agent its own identity, so its actions are not mixed up with a person's or another agent's.

Your next step

If you want your agents' access locked down, book a free 30-minute 1:1 assessment. I look at your setup with you and tell you what the work would take.

Drafted with AI help, edited and approved by me.

New articles by email

I will never sell your details, for any reason, and I never send spam.

How I use your details
  • I, Abhijit Chatterjee Chattopadhyay of The Logic Atelier, hold them, in my own newsletter system hosted on Fly.io. Emails go out through Resend.
  • Your name is for greeting you, your email is for sending new articles, and your phone number, if you give it, is so I can reach you directly about my work.
  • If you do not confirm your email within 7 days, I delete the sign-up. If you unsubscribe, I delete your details 30 days later.
  • If Google Analytics is running on your visit, it is told that a sign-up was made, and nothing you typed.
  • To see or delete what I hold about you at any time, email hello@thelogicatelier.com.
  • The full detail is on the privacy page.
NextThe small model that beat the LLM
← cd ../writing

Your AI agent has more access than you think

agent-access.mdfront matter
title: Your AI agent has more access than you think
date: 2026-10-21
reading_time: 2 min
author: Abhijit Chatterjee Chattopadhyay
agent_access_cover_03e961b0_d8900689b2.pngcover
The Logic Atelier logo above the words: Your AI agent has more access than you think.

Your AI agent was set up to do one job. It can probably do far more.

The role nobody meant to grant

Agents get access the quick way. Someone needs it working today, so the agent gets an existing role, an admin key or a service account that already has what it needs, and a lot more.

AWS describes the trap in the same Trust in Depth (opens in a new tab) guidance I quoted on Monday: "Grant an admin role for one task, and the agent can inherit permissions across your whole environment." It also warns about "capabilities no one granted on purpose, the ones bundled with a tool or a role."

The agent does not need to be malicious for this to hurt. A wrong step, a confusing instruction or a prompt injection is enough, if the access is there.

Why it matters more for agents

A person with too much access still has to decide to use it. As AWS notes in the same piece, an agent finds its own path to finish a task, picking tools and steps no developer scripted. Whatever it can reach is on the table. And unlike a person, an agent often has no identity of its own, so nobody can tell afterwards which one acted.

So scope the agent before it runs. AWS calls scoping "where to invest the most effort. Problems caught here are cheap to fix."

Three checks before an agent goes live

  1. What is its job, in one sentence? "Answer delivery questions for existing orders." Not "help customers".
  2. What does that job need? List the data, tools and actions. Read orders, yes. Issue refunds, no.
  3. What did it get anyway? Compare that list with the access it actually holds. Everything extra comes off, including what arrived bundled with a tool or a role.

The gap between check 2 and check 3 is your real risk.

Do this today

  • Pick your most-used agent and write its job in one sentence.
  • Pull the list of permissions its account or role actually has.
  • Remove anything the sentence does not need.
  • Give each agent its own identity, so its actions are not mixed up with a person's or another agent's.

Your next step

If you want your agents' access locked down, book a free 30-minute 1:1 assessment. I look at your setup with you and tell you what the work would take.

Drafted with AI help, edited and approved by me.

subscribeemail

// I will never sell your details, for any reason, and I never send spam.

How I use your details
  • I, Abhijit Chatterjee Chattopadhyay of The Logic Atelier, hold them, in my own newsletter system hosted on Fly.io. Emails go out through Resend.
  • Your name is for greeting you, your email is for sending new articles, and your phone number, if you give it, is so I can reach you directly about my work.
  • If you do not confirm your email within 7 days, I delete the sign-up. If you unsubscribe, I delete your details 30 days later.
  • If Google Analytics is running on your visit, it is told that a sign-up was made, and nothing you typed.
  • To see or delete what I hold about you at any time, email hello@thelogicatelier.com.
  • The full detail is on the privacy page.

vi.Contact

06 · contact

Talk to me.

Talk to me.

Production AI and software architecture. Based in Dubai, working with teams in the Emirates and Ireland.

I work Dubai hours, Monday to Friday, the same week you do. Dubai is three hours ahead of Dublin in summer and four in winter, so my afternoon is your morning.

I will never sell your details, for any reason, and I never send spam.

How I use your enquiry
  • Your enquiry comes to me, Abhijit Chatterjee Chattopadhyay of The Logic Atelier. The form is handled by Formspree, which passes it to my inbox and keeps a copy in my Formspree account.
  • I use it to read your enquiry and reply. If no work follows, I delete it within 12 months, and sooner whenever you ask.
  • If Google Analytics is running on your visit, it is told that an enquiry was sent, and nothing you typed.
  • To see or delete what I hold about you, email hello@thelogicatelier.com.
  • The full detail is on the privacy page.