A figure seen from behind, the back of the head showing, sorts a heaped basket of envelopes, wax recording cylinders and loose papers into three places on a desk, a bound rulebook on a lectern, a slim folder for one task and an overflowing tray, while a small robot at the desk's end reads only the rulebook, the whole scene drawn with wide empty margins on every side.

Nielsen Norman Group

The 3 Roles of Context for AI Agents

Tanner Kohler · 18 September 2026

Takeaway

A Nielsen Norman Group study found that heavy users of the AI agent Claude sort its background information into three kinds, including standing rules on what needs their approval.

Summary

Tanner Kohler of Nielsen Norman Group, a UX research and training firm, studied people who use the AI agent Claude daily for substantial professional work. Every participant had built a library of files, databases and live connections for Claude to draw on. Kohler found that this information plays three roles: standing guidance for every task, material for one project, and raw streams such as email and meeting transcripts.

Key points

Global information rarely changes and applies to every task, such as templates, brand rules and personal preferences. Local information belongs to one task or project, such as a to-do list.
Ambient information is the raw stream of email, meeting transcripts, chat messages and analytics. Participants made no effort to sort it and let Claude search through it.
The global files included rules on which actions needed the user's approval. For one participant, Claude needed permission for anything public-facing.
Kohler advises anyone building such a library, for themselves or for a team, to decide what role each piece plays. The study observed individuals, not teams.

Implication

A team that shares one set of standing instructions for its agents has to decide who may change them, because every task follows them. This applies a study of individuals to teams and is an inference, not a finding of the study.

Suggested actions

Product

Write down, in the standing instructions an AI agent reads before every task, which actions it may take on its own and which need a person's approval.

Design

List in an AI agent's standing instructions the design rules it must follow and the design changes that need a designer's approval.

Derived by Working Surface from the article. Source line: Governance rules: Which actions required the user's approval, and which Claude could take on its own.

Source issue

20 September 2026: the standing instructions people give AI agents, including what needs approval and what the agent must ask first.