For leaders, builders, compliance teams and newcomers

AI governance, from first step to full program.

Learn the basics, find the rules for your regions, assess your AI systems, and produce the documents and evidence you need. Free, private, and nothing leaves your device.

Start here Find your rules

Your clear path through AI governance, region by region.

Last reviewed

3 Oct 2026

Regions covered

EU, US, UK, India, plus global standards

Privacy

No accounts, no cookies

Status

For reference only, not legal advice

Audiences

Where do you want to start?

I lead a business

Understand risk, accountability and the questions to ask your teams.

Top tools: Readiness check, Start here planner

Start here →

I build AI products

Get practical guidance on controls, testing and documentation.

Top tools: Screening check, Evidence checklist

Start here →

I work in compliance or legal

Map laws and standards to concrete obligations and evidence.

Top tools: Compare regions, Crosswalk

Start here →

I buy AI products

Ask suppliers the right questions and get the answers into the contract.

Top tools: Lifecycle: design controls, Templates

Start here →

I work in HR

Use AI in hiring and people decisions fairly, with the right notices.

Top tools: Compare regions, Screening check

Start here →

I'm new to AI governance

Learn the basics in plain language, with no jargon and no prior knowledge.

Top tools: Glossary, Quiz

Start here →

The lifecycle

Eight stages, one clear output each.

Every page, tool and template belongs to one or more of these stages.

Set up and assess

Stage 1: Set up governance

Decide how AI is governed and who decides.

Output: AI policy, governance charter, RACI, training plan

Stage 2: Inventory and intake

Know every AI system and control new ones at the door.

Output: AI register, intake records

Stage 3: Assess risk and impact

Understand how risky each system is and for whom.

Output: Screening result, applicability record, impact assessment

Stage 4: Design controls and document

Choose and document the right safeguards.

Output: Control plan, model or system card, data documentation, vendor file

Build, run and improve

Stage 5: Test and validate

Show it works fairly, accurately and securely.

Output: Test plan, test and bias reports, red-team record

Stage 6: Approve and deploy

Make an informed go-live decision with people informed.

Output: Approval record, transparency notices, oversight procedure, instructions

Stage 7: Monitor and respond

Catch problems early and respond well.

Output: Monitoring reports, incident log, notifications

Stage 8: Review, report and retire

Keep improving, report honestly, retire safely.

Output: Periodic reviews, board reports, decommissioning record

See the lifecycle · Not sure where you are? Take the readiness check.

Regions

The rules, region by region.

EU

European Union

AI-specific binding law? Yes

Risk-based AI regulation, layered on data protection and product law.

Next key date: 2 Dec 2026: New prohibited practices

Open European Union →US

United States

AI-specific binding law? No comprehensive federal AI statute. Yes in some states.

Executive action and agency enforcement of existing laws, plus a growing body of state law.

Next key date: 1 Jan 2027: Several state laws take effect

Open United States →UK

United Kingdom

AI-specific binding law? No

Principles-based and regulator-led. Existing laws and sector regulators apply AI principles in their areas.

Next key date: Watch: Regulating for Growth Bill and any AI bill

Open United Kingdom →IN

India

AI-specific binding law? Partly

Voluntary national guidelines, with binding duties through IT rules and data protection law, plus sector regulators.

Next key date: 13 Nov 2026: DPDP: consent manager provisions

Open India →Global

Global standards

AI-specific binding law? No (voluntary or non-binding)

Voluntary frameworks, certifiable standards and intergovernmental principles you can use as a backbone for any region.

Next key date: Watch: NIST AI RMF revision

Open Global standards →

Upcoming deadlines

The next five dates across your regions.

13 Nov 2026
Upcoming
IN DPDP: consent manager provisions
Consent manager provisions of the DPDP Rules apply from 13 Nov 2026 (12 months after notification).
In 41 days
2 Dec 2026
Upcoming
EU New prohibited practices
Two prohibitions added by the Digital Omnibus apply: using AI to generate non-consensual intimate imagery of identifiable people, and child sexual abuse material.
In 60 days
1 Jan 2027
Upcoming
US Several state laws take effect
Colorado SB 26-189, California automated decision obligations, and the New York RAISE Act and Illinois SB 315 frontier laws take effect from 1 Jan 2027.
In 90 days
13 May 2027
Upcoming
IN DPDP: most duties
Most remaining DPDP duties (notice and consent, security safeguards, breach intimation, retention and Significant Data Fiduciary obligations) apply from 13 May 2027 (18 months after notification).
In 222 days
2 Dec 2027
Upcoming
EU High-risk, stand-alone
Annex III uses such as employment, credit and education.
In 425 days

Dates are a summary. Confirm them with qualified counsel for your own case. See the full calendar.

Tools

32 tools, grouped by stage.

Everything runs in your browser.

Stage 1

Score your program, then pick the next move.

Rate five areas honestly. The results update as you go. Open the full check.

Policy and roles

Is there an approved AI policy with a named owner?

System inventory

Is every AI system, including vendor tools, on a register?

Risk classification

Is each system placed in a risk tier using written criteria?

Controls and testing

Are testing, documentation and oversight matched to each tier?

Monitoring and audit

Are logs, incidents and changes tracked and reviewed?

Your score

0%

Policy and rolesNot started

System inventoryNot started

Risk classificationNot started

Controls and testingNot started

Monitoring and auditNot started

Focus next

Open the lifecycle stage

Your answers stay in this browser tab. Screening result only, not a compliance rating.

Updates

What changed recently.

3 Oct 2026 · Global

Key EU and US dates corroborated

Publication and entry-into-force of Regulation (EU) 2026/1744, the 2 Dec 2027 Annex III date, the 2 Dec 2026 marking transition and new prohibitions, the AI literacy wording change, EO 14365, Colorado SB 26-189, Illinois SB 315 and New York RAISE were checked against secondary legal summaries. Primary texts were not read, so nothing is marked confirmed.

See the section →

3 Oct 2026 · IN

India dates made exact; UK commencement noted

DPDP phase dates are now shown as 13 Nov 2026 and 13 May 2027, the IT Rules takedown deadlines are summarised, and the UK automated decision-making reforms are noted as in force from 5 Feb 2026. All from secondary summaries or GOV.UK guidance; primary Rules and commencement regulations not read.

See the section →

2 Oct 2026 · Global

Initial publication

First release (phase 1). EU AI Act dates reflect the Digital Omnibus on AI, in force since 27 July 2026. Primary-source verification of region content is still to be completed; see the verification report.

See the section →

All updates · Subscribe via RSS