Skip to content
Back to Blog
AI literacy series
May 13, 2026
·by Piyush·4 min read

Trusting AI at Work: Approvals, Boundaries, and Receipts

ContextOS
AI Literacy
Trust
Governance
Agents
Share:XHN

Trusting AI does not mean believing it is always right.

Trusting AI means knowing:

  • what it can do,
  • what it cannot do,
  • when it must ask,
  • what evidence it used,
  • who approved the action,
  • how to inspect what happened.

That is a very different kind of trust.

The bank analogy

A bank does not trust employees by giving everyone vault access.

It creates roles:

  • teller,
  • supervisor,
  • auditor,
  • fraud analyst,
  • branch manager.

Each role can do different things. Some actions need approval. Some require evidence. Everything leaves a record.

AI agents need the same structure.

Five levels of authority

Use this simple ladder:

LevelPlain EnglishExample
LookAI reads or summarizes”Find the order”
DraftAI prepares but does not send”Draft the email”
Ask outsideAI calls an external service”Check shipping status”
Act for userAI does something on delegated authority”Schedule the meeting”
High-impact actAI changes money, access, legal state, or sensitive data”Issue refund”

In ContextOS, these are approval modes: read_only, local_write, network, delegated, and destructive.

Non-technical leaders can use this ladder in meetings. It makes risk concrete.

Ask: what should require a gate?

A gate is a moment where AI must pause and ask.

Gate examples:

SituationWhy gate?
Money movesFinancial risk
Customer receives messageRelationship risk
Account access changesSecurity risk
Employee data usedPrivacy risk
Legal or compliance claim madeRegulatory risk
Ambiguous policyJudgment risk
Missing evidenceAccuracy risk

The gate is not a failure. It is part of the system.

What an approval should show

A useful approval request includes:

  • proposed action,
  • reason for action,
  • evidence used,
  • policy or rule,
  • possible side effect,
  • what happens if rejected,
  • who is approving,
  • final receipt.

Bad approval:

Approve refund?

Good approval:

Approve INR 9,000 refund for Order 123. Evidence: identity verified, order delivered late by 6 days, refund policy section 4.2 applies, amount exceeds self-serve threshold. If approved, payment system will issue refund once with idempotency key R-912.

The second version lets a human take responsibility.

Receipts matter

After AI does important work, it should leave a receipt.

ContextOS calls this a DecisionRecord.

A receipt answers:

Receipt questionWhy it matters
What work was requested?Scope
What evidence was used?Grounding
What tools were called?Action trace
What policy applied?Governance
Who approved?Accountability
What changed?Impact
Can we replay it?Audit and learning

If the system cannot produce a receipt, it should not perform important work.

Boundaries are not bureaucracy

Boundaries make AI useful.

Without boundaries, people either overtrust the system or refuse to use it.

With boundaries, people know:

  • this AI can summarize but not send,
  • this AI can draft but not approve,
  • this AI can recommend but not execute,
  • this AI can execute only under this threshold,
  • this AI must escalate these cases.

Clear boundaries increase adoption because users understand the tool.

Trust checklist for business teams

Before launching AI into a workflow, ask:

  1. What can it read?
  2. What can it draft?
  3. What can it send or change?
  4. What requires approval?
  5. What evidence is mandatory?
  6. What must it refuse?
  7. What receipt does it leave?
  8. Who reviews mistakes?
  9. How can we stop or roll back the system?

These questions are enough to find most hidden risks.

Common misunderstanding

People often think the choice is:

Fully automate or do nothing.

The better choice is:

Decide which parts can be assisted, drafted, delegated, or gated.

Most good AI systems start by helping humans do better work. They earn more authority over time.

Found this useful? Share it.

Share:XHN
Analytics consent

We use Google Analytics to understand site usage. You can opt in or decline.