For law firms
Claude, implemented and managed for your law firm.
Your associates are already drafting, summarizing, and researching with AI tools — some of them on personal accounts, with client matter facts in the prompt. The duty of confidentiality did not change when the tools did.
Already happening
What your staff are doing today.
Not to assign blame. The assessment starts by finding it, tool by tool, so the ungoverned path can be replaced with a governed one.
- —Drafting correspondence and first-pass documents in a consumer chatbot
- —Summarizing discovery, depositions, and contracts by pasting them in
- —Research and cite-checking with tools that were never told what a privileged document is
What already binds you
The rules did not change when the tools did.
Every finding in the report is tied to the duty or standard it falls under, so the fix is defensible to whoever asks.
- —The duty of confidentiality and the attorney-client privilege, for every matter
- —Your state bar's guidance on generative AI, and the ABA's Formal Opinion 512 on the same
- —The duty of competence and supervision — work that goes out under your name is reviewed by you
The assessment
Two weeks. Fixed fee. You keep the report.
Two weeks inside your law firm. We find where AI is already in use, which client information it touches, and under what policy — interviews with the people doing the work, a review of the tools they actually use, and a walk through the policies you already have.
A baseline Company Skoor — your score against the published rubric — and a written finding for every gap, tied to the duty or standard it falls under, with a remediation plan priced for phase two. Yours to keep whether or not you continue.
A fixed fee, quoted on your size before you commit and fixed once quoted. If you go on to implementation, the full fee is credited against it.
If you continue: Claude on your own devices, a written profile per role, connectors allowlisted, the audit trail wired to a store you control, staff trained. Then annual maintenance, with every agent scored continuously.
Rubric v1.0 is scoped to CPA firms. A rubric for law firms is published here before the first engagement, and the standards it tracks are named in the report. The rubric →
What a deployment looks like
A written profile per role.
Permissions scoped to what that role may touch, connectors allowlisted, review where the work goes out under your name. Examples, not a menu — yours are written from the findings.
- —Associate — drafting and research, no client identifiers in unapproved tools
- —Paralegal — document intake and summaries, matter-scoped connectors only
- —Billing and intake — client data with consent, retention where the firm controls it
Why the number is worth something
Start with the assessment.
Two weeks. A fixed fee, quoted on your size and credited in full against implementation. A baseline Company Skoor and a written finding for every gap — yours to keep either way.
Also for consulting firms · medical and dental practices · accounting firms
Claude is a product of Anthropic, PBC. Skoor is an independent implementer and is not affiliated with or endorsed by Anthropic. A Skoor scores controls; it is not a legal opinion.