Book a call

Building Answerable AI for Enterprise Leaders.

Know the rules. Check your AI. Prove it with evidence.

Start with a free ten-minute check. When you need more, we scan your AI agents, map them to the laws you are held to, and give you an evidence package an auditor or regulator can verify.

Your AI Governance Index

72 / 100

Illustrative sample. Not a client result.

On the name

Calibre: the standard by which quality is measured. Not caution. Not delay. A deliberate return to what can be proven, before the decision is made.

From the founder’s note →

Start where you are

What brings you here today?

Pick the sentence closest to your situation. Each one takes you to the right first step, most of them free.

A deadline is coming

An AI law or regulator date is close and you need to know what applies to you.

See the deadline tracker →

Our agents worry us

AI agents and copilots are connected to email, files and systems, and nobody can say what they can reach.

Run the Agent Exposure Check →

An audit is coming

A board, customer, auditor or regulator has asked you to prove your AI governance.

Take the Proof Test →

We want it running

You want discovery, scoring and evidence every month, not once a year.

See the platform →

How customers work with us

Four steps, and you decide how far to go.

  1. 1. Learn

    Free

    Plain-language guides to AI laws in each region, with primary sources and a date checked.

  2. 2. Check

    Free, 10 minutes

    Self-service tools that run in your browser. Nothing you type is sent to us.

  3. 3. Scan

    One-time fee

    We scan your AI agents across platforms and deliver a scored report and a fix list within ten working days.

  4. 4. Run

    Monthly plan

    Your own Fabric workspace keeps discovering, scoring and recording evidence, with advisory time when you need it.

See pricing →

The governance gap

Most organisations have AI policies. Few can show they work.

Boards ask for assurance, regulators ask for proof, and teams are left with spreadsheets and slide decks.

No single number

Governance maturity is described in paragraphs, so it cannot be tracked, compared or reported to a board.

Claims without evidence

Controls exist on paper. When an examiner asks for proof, teams scramble to assemble it.

No clear next step

Gaps are known in general, but not ranked by impact, so effort goes to the loudest issue.

Built around the frameworks your board and regulators ask aboutISO/IEC 42001NIST AI RMFEU AI ActCERT-InGulf and Asia-Pacific rules

The Answerable AI Toolkit

Free tools. Each answers one hard question in about ten minutes.

Free · live

Agent Exposure Check

What could your AI agent do if it were tricked?

Select the permissions of an AI agent or MCP server. See which actions an attacker could reach through prompt injection, mapped to OWASP LLM risks and MITRE ATLAS techniques.

For: CISOs, AI platform teams

Open the tool
Free · live

AI Act Classifier

Is your AI system prohibited, high-risk or limited-risk?

Answer plain questions about one AI system. Get its likely EU AI Act category, your role, and the obligations and dates that follow, with an Asia-Pacific overlay.

For: Compliance, legal, product owners

Open the tool
Free · live

India AI Audit Tier Calculator

Which CERT-In audit tier does your AI system fall into?

Score one AI system on the eight factors in CERT-In's AI Security Auditing Guidelines. Get its tier, the audit depth that follows, and the escalation triggers to watch.

For: Indian banks, fintechs, government suppliers

Open the tool
Free · live

Gulf AI Go-Live Navigator

Which Gulf rules apply before your AI goes live?

Pick your jurisdictions across the UAE, DIFC, ADGM, Saudi Arabia, Qatar, Oman, Bahrain and Kuwait. Get the rulebooks that apply and a pre-launch checklist.

For: Gulf banks, insurers, government entities

Open the tool
Free · live

Proof Test

Could you prove your AI governance within 48 hours?

Answer the 12 evidence requests an auditor would make. Get a readiness score by ISO/IEC 42001 clause, NIST AI RMF function and EU AI Act article.

For: Risk, internal audit, AI governance leads

Open the tool

The tools run entirely in your browser. No sign-up, no tracking, and nothing you enter leaves your device.

Know the rules

AI law, explained for where you operate.

Each guide states who the law applies to, what it requires, the dates, and links to the official text. Every page shows when we last checked it.

European Union

AI Act, Article 50 transparency, GDPR and DORA.

India

DPDP Rules, CERT-In AI audit tiers, RBI FREE-AI.

Australia

Privacy Act automated decisions, CPS 230, AI safety rules.

Gulf states

UAE, DIFC, ADGM, Saudi Arabia, Qatar and Oman.

United States

Colorado, California ADMT and state-by-state rules.

Standards

ISO/IEC 42001, ISO/IEC 27001, NIST AI RMF, OWASP.

Open the global deadline tracker →

Calibre Governance Fabric

One fabric across your AI estate, from inputs to verifiable evidence.

Bring together your agents, use cases, policies and controls. The Fabric scores them, records the evidence and shows what to fix first.

Your AI estate

  • AI agents and copilots, found across platforms
  • AI use-case register
  • Policies, controls and owners
  • Models, vendors and incidents

Governance Fabric

Discover. Score. Evidence.

Each organisation gets its own workspace and its own separate database, with role-based access for assessors and reviewers. Connectors are read-only.

What you get

  • AI Governance Index
  • Verifiable evidence package
  • Prioritised gap roadmap
  • Board-ready summary

What the Fabric does

Three things, held together as one discipline.

Measure

A transparent AI Governance Index across the dimensions that matter, scored the same way every time.

Track movement quarter on quarter, compare business units, and show the board a number with the reasoning behind it.

Evidence

Every score links to evidence, packaged so a third party can check it independently.

Reviewers can use the public verification page to confirm an evidence package without needing an account.

Verify an evidence package →

Improve

A ranked roadmap that shows which gaps move the Index most, and who owns each fix.

Spend effort where it changes your risk position, not where it is easiest to report.

Who it is for

Built for the people who must answer for AI.

Boards and CXOs

A clear number, a trend and a short list of decisions, without wading through technical detail.

Risk and compliance

Evidence you can hand to an auditor, mapped to the frameworks you are held to.

AI and technology leads

Know which controls are missing, so governance helps teams ship rather than slowing them.

Advisory + Practitioners + Platform = Answerable AI

How an engagement runs

From first conversation to a defensible baseline.

1. Scope

A 30-minute call to agree the AI estate, frameworks and people in scope. A fixed quote follows.

2. Connect

We send a one-page access guide. Your team grants read-only access; nothing is changed in your systems.

3. Score

The Fabric finds your agents, scores them and maps each gap to the rules you are held to.

4. Prove

We walk you through the report and hand over a verifiable evidence package and a 90-day plan.

First findings within 10 working days of read-only access being granted.

Our service promise

What you can expect from us, in writing.

One business day

Every enquiry gets a reply from a person, not a bot, within one business day.

A named lead

One accountable person from scoping call to closure. You never repeat yourself.

No surprises

Fixed quotes, no hidden extras, and a written change request before any scope change.

Your data, your control

Read-only access, separate database per client, deletion certificate at the end.

A note from the founder

[FOUNDER NOTE: two or three sentences on why you built CalibreInsight. Suggested close: Governance plus evidence equals answerable AI. The reason a customer stays is the reason a regulator nods: the work holds up. If that is your standard, write to us.]

Chirag Dani · Founder, CalibreInsight

Latest perspective

Thirteen new guides to AI law are being published

Gulf, India, Australia, the United States and the European Union, each with primary sources and a free tool to go with it.

Browse insights →

Contact us

If you are building AI that must be answerable and defensible, we would like to meet you.

We take a focused number of engagements. Each begins with a conversation about what you need to prove, and to whom.

This opens your own email app with the message filled in. Nothing is sent to our servers until you press send there.