Sample This is a fictional example. Halstead & Rowe Solicitors is a firm we invented to show the dashboard; it is not a client. It is drawn by the same system as a real client's page. The buttons are switched off, and parts of the framework itself are shown only on a client's own dashboard.

AI compliance dashboard · UK + EU Combined Compliance Framework / Home

Download compliance report Log an incident Add AI tool

Halstead & Rowe Solicitors

12 things need you, 1 overdue. Your cover runs to 14 August 2027.

Deadlines that reach you

All dates →
11days

2 December 2026. Two prohibited practices the Digital Omnibus added — AI that makes intimate images of a real person without consent, and AI that makes child sexual abuse material (Article 5(1)(ba) and (bb)) — the other eight prohibitions already apply; and the deadline for generative AI tools already on the market before the transparency duties applied to mark the content they make (Article 50(2)).

2 Aug 2026EU AI Act
2 Dec 2026EU AI Act
2 Aug 2027EU AI Act
2 Dec 2027EU AI Act
Today

Something changed?

Log it and your record stays current

Every AI tool against every check

4 tools · Open AI tools →
CheckChatGPT TeamLexis+ AIMicrosoft 365 CopilotOtter.ai
Covered by your documentsNamed in the framework we wrote youCoveredCoveredCoveredNot yet
Risk ratedUnacceptable, high, limited or minimalHighLimitedLimitedNot yet
Reviewed on timeThe riskier the tool, the soonerDue soonIn dateIn dateOverdue
Someone looks after itA named ownerNamedNamedNamedNamed
No open incidentLogged and closedNoneNoneNoneOpen

Needs you

12 open · 1 overdue

Activity trail

All of it →
  • 20 Nov 2026, 16:41Added Otter.ai to the AI toolsOn the dashboard
  • 18 Nov 2026, 13:24Emailed Tom Ellis the AI policy to read and signOn the dashboard
  • 15 Nov 2026, 10:07Logged an incident: A client call was recorded and transcribed before the client was told.On the dashboard
  • 7 Nov 2026, 15:50Added Tom Ellis to the staff listOn the dashboard
  • 13 Oct 2026, 13:33Closed an incident: A draft letter cited a case that does not exist.On the dashboard

Policy library

All 3 →
UK GDPR Gap Analysis for AI Tools AI Acceptable Use Policy Employee AI Guidelines — Practical Dos and Don'ts
3 of 4 staff have signed the AI policyPeople

Your AI tools

4 on record

The riskier the tool, the sooner it is looked at again: high risk every 3 months, limited every 6, minimal once a year.

ChatGPT Team

High risk

Used for: Drafting client letters; summarising case bundles

The risk: Client files hold confidential and sensitive personal data. Used only on the business plan with training switched off, and a solicitor checks every draft before it leaves the firm.

Owner
Practice manager
Review
Review due 1 December 2026
We switched to another tool

ChatGPT Team comes out of your documents and the new tool goes in. We update them overnight and email you the new set.

Lexis+ AI

Limited risk

Used for: Legal research

The risk: A legal research tool with confidentiality terms in the subscription. Every authority it suggests is checked at source before it is cited.

Owner
Head of litigation
Review
Next review 9 February 2027
We switched to another tool

Lexis+ AI comes out of your documents and the new tool goes in. We update them overnight and email you the new set.

Microsoft 365 Copilot

Limited risk

Used for: Emails, meeting notes and first drafts

The risk: Stays inside the firm’s Microsoft account under its existing data terms. Drafts are checked for accuracy before they are relied on.

Owner
IT partner
Review
Next review 9 February 2027
We switched to another tool

Microsoft 365 Copilot comes out of your documents and the new tool goes in. We update them overnight and email you the new set.

Otter.ai

Not yet rated

Used for: Notes from client calls

Not covered by your documents yet. We update them overnight and email you the new set.

Owner
Practice manager
Review
Review overdue
We switched to another tool

Otter.ai comes out of your documents and the new tool goes in. We update them overnight and email you the new set.

Add an AI tool

Started using a new AI tool?

Add it here. Overnight we rate its risk, update your documents to cover it and email you the new set.

Replacing a tool you already use? Press “We switched to another tool” on it instead, so the old one comes out of your documents.

What adding a tool changes

1
It goes on your record todayDated in your history, with who looks after it.
2
We rate its riskUnacceptable, high, limited or minimal. Until then it shows as not yet rated.
3
It gets a review dateHigh risk every 3 months, limited every 6, minimal every 12.
4
Your documents are rewritten to cover itWe update them overnight and email you the new set.
5
It appears on your compliance statementWith its risk level, once rated.

Your staff

4 on record

Everyone who uses AI at work: when they signed your AI policy, and when they were last shown how to use AI safely. Add a work email and we send each person the policy to read and sign with their own name — the strongest record you can hold. Training is refreshed every 12 months.

PersonAI policyAI training
Ruth HalsteadPartner Signed16 August 2026 Trained16 August 2026 · Firm AI briefing (45 minutes)
Daniel RowePartner Signed16 August 2026 Trained16 August 2026 · Firm AI briefing (45 minutes)
Priya ShahSolicitor Signed it themselves23 September 2026 Not trained yet
Tom EllisTrainee Asked to signEmailed 18 November 2026 · not signed yet Not trained yet
Add your whole team at once

One person per line: name, then role and work email if you have them. You can paste a column straight from a spreadsheet or your HR system. Someone already on the list is skipped.

Incidents

1 open

When an AI tool gets something wrong — or nearly does — write down what happened and what you did. Near misses count: they are the proof your checks work.

Open

A client call was recorded and transcribed before the client was told.

15 November 2026 · Otter.ai

What we did: Recording deleted; client told and apologised to. Waiting on a script for opening calls.

Client letter and file note

Worded from what you logged, for you to finish and send from your own email. Fill in the square brackets. Nothing is sent from here, and changes you make here are not saved.

Closed 14 Oct 2026

A draft letter cited a case that does not exist. Caught at the solicitor check before it was sent.

11 October 2026 · ChatGPT Team

What we did: Draft corrected. Reminder to all fee earners: every authority checked at source.

Client letter and file note

Worded from what you logged, for you to finish and send from your own email. Fill in the square brackets. Nothing is sent from here, and changes you make here are not saved.

Log an incident

Describe what happened, not who it happened to: leave out your clients' and customers' names and details.

Your documents

PDF and editable Word

Every document in your framework, as a PDF and an editable Word file, whenever you need one.

Your whole framework PDFWord
UK GDPR Gap Analysis for AI Tools PDFWord
AI Acceptable Use Policy PDFWord
Employee AI Guidelines — Practical Dos and Don'ts PDFWord

35 more documents in this framework — listed and downloadable on a client's own dashboard

Version history

0 updates

Each time we rewrote your documents, and why. The set above is always the current one.

No updates yet. These are the documents as we first delivered them.

History

Latest 14

Every change to your dashboard, dated, and how it was made. It is your record that the policy is followed through the year, not just signed once.

  • 20 Nov 2026, 16:41 Added Otter.ai to the AI tools On the dashboard
  • 18 Nov 2026, 13:24 Emailed Tom Ellis the AI policy to read and sign On the dashboard
  • 15 Nov 2026, 10:07 Logged an incident: A client call was recorded and transcribed before the client was told. On the dashboard
  • 7 Nov 2026, 15:50 Added Tom Ellis to the staff list On the dashboard
  • 13 Oct 2026, 13:33 Closed an incident: A draft letter cited a case that does not exist. On the dashboard
  • 11 Oct 2026, 10:16 Logged an incident: A draft letter cited a case that does not exist. On the dashboard
  • 22 Sep 2026, 15:59 Priya Shah signed the AI policy (typed name: Priya Shah) By email
  • 2 Sep 2026, 12:42 Reviewed ChatGPT Team — Practice manager On the dashboard
  • 15 Aug 2026, 17:25 Daniel Rowe trained to use AI safely — Firm AI briefing (45 minutes) On the dashboard
  • 15 Aug 2026, 14:08 Ruth Halstead trained to use AI safely — Firm AI briefing (45 minutes) On the dashboard
  • 15 Aug 2026, 11:51 Daniel Rowe signed the AI policy On the dashboard
  • 15 Aug 2026, 16:34 Ruth Halstead signed the AI policy On the dashboard
  • 13 Aug 2026, 13:17 Reviewed Microsoft 365 Copilot — IT partner On the dashboard
  • 13 Aug 2026, 10:00 Reviewed Lexis+ AI — Head of litigation On the dashboard

What changes for you

The next dates in the laws your framework covers, each checked against the law itself. When a law changes, we update your documents and tell you what moved.

2Dec2026 Two prohibited practices the Digital Omnibus added — AI that makes intimate images of a real person without consent, and AI that makes child sexual abuse material (Article 5(1)(ba) and (bb)) — the other eight prohibitions already apply; and the deadline for generative AI tools already on the market before the transparency duties applied to mark the content they make (Article 50(2)).EU AI Act The law ↗
2Aug2027 Every Member State must have an AI regulatory sandbox running.EU AI Act The law ↗
2Dec2027 High-risk obligations for Annex III systems — including AI used in recruitment, credit scoring, education and essential services.EU AI Act The law ↗
2Aug2028 High-risk obligations for Annex I systems — AI embedded in products already covered by EU product safety law.EU AI Act The law ↗

Your proof

3 of 9 in date

Each promise in your AI policy needs a record that proves you keep it. Say where each one is held and who owns it; each needs confirming again when it falls due.

  • In date Approved people and tools listKept in: AI policy, section 2 — practice handbook · Practice manager Confirm by 10 January 2027
  • In date Tool request logKept in: Approved tools list — shared drive / Compliance · IT partner Confirm by 8 October 2027
  • In date Staff sign-off formsKept in: Signed staff acknowledgements — HR folder · Practice manager Confirm by 11 December 2026
  • Out of date Output check logKept in: Draft-check log — matter files · Head of litigation Confirm by 13 October 2026
  • Not recorded Human decision points
  • 4 more records your policy requires — listed on a client's own dashboard

Update your proof

Your compliance statement

A one-page statement to send a client, insurer or supplier who asks how you manage AI: your tools and their risk levels, the documents you hold and the standards they are aligned to. No names or personal details. It stays valid while your cover runs.

Every Complete framework, and the UK + EU Combined framework, comes with this dashboard — built from your own documents, kept current all year.

See the frameworks →

Pick your market. Buy today.

Fixed price. Plain English. Written for how your business actually uses AI.

United Kingdom

UK AI Compliance — Foundation

£797/year

Buy now → What's included

European Union

EU AI Act Compliance — Foundation

€1,197/year

Buy now → What's included

United States

US AI Compliance — Foundation

$997/year

Buy now → What's included

Canada

Canada AI Compliance — Foundation

$997/year

Buy now → What's included

Middle East

Middle East AI Compliance — Foundation

$997/year

Buy now → What's included

Far East

Far East AI Compliance — Foundation

$997/year

Buy now → What's included

Australia and New Zealand

Australasia AI Compliance — Foundation

A$1,597/year

Buy now → What's included

Not sure which laws reach you? Take the free two-minute check and it will tell you, or ask us on WhatsApp.

Call Now See prices