Human-in-the-loop, every output
Every AI-generated document, draft, advice memo, billing narrative, or research result is labelled AI-generated until a fee earner explicitly reviews and approves. Nothing reaches clients until that sign-off.
Responsible AI
Governed AI for law firms — assistive, audited, and human-reviewed.
CounselOS embeds AI inside your practice workflow: drafting, research memos, contract review flags, and billing narratives. Every output is labelled, logged, and held for fee earner approval before it reaches a client or court file. CounselOS does not provide legal advice — your lawyers remain accountable.
That discipline is designed to support professional conduct obligations: documented policy, protected client details, and a clear audit trail when regulators or clients ask how AI was used.
Six non-negotiables.
Client details protected before outside AI
Names and identifying details are shielded before anything is sent to an outside AI service. Your firm's mapping stays on our servers — there is no setting to turn this off.
Hallucination disclosure on every output
Every AI-generated research memo and contract review carries the disclaimer "Verify all legal citations independently before relying on them." Plus warnings on any citation we could not verify.
Audit trail — every use recorded
Every AI assist is logged with who triggered it, what feature was used, which matter it related to, and the outcome. Exportable for regulatory review or internal audit.
Citation verification
Research memos include legal citations. CounselOS checks what it can and flags references you should verify before relying on them. Legal research uses governed AI with citation checking. Connections to paid publishers (such as Lexis or Jade) are only available when your firm has a signed arrangement — website demos may use sample cases, not a live publisher feed.
Categorically blocked data types
Mandatory reports, child protection orders, domestic violence orders, and national security data never reach an external AI provider — even in de-identified form. Refused at the gateway.
Your firm's AI policy. Generated and dated.
CounselOS generates a firm-specific AI usage policy on day one, tailored to your state, practice areas, and risk appetite. Stored as a Document. Refreshed annually. Versioned in the audit log.
A firm with documented, current AI governance is better placed in a professional conduct review than a firm whose lawyers use unapproved tools in browser tabs without records.
See a sample firm AI policySAMPLE EXCERPT
"Bridgewater Legal Pty Ltd ("the firm") uses CounselOS as its sole approved AI-assisted practice management platform. The firm's principals retain responsibility for the accuracy, completeness, and ethical use of every AI-generated output. No client confidential information is to be entered into any AI tool other than CounselOS without the express written approval of the principal solicitor…"
Last reviewed: 16 May 2026 · Next review due: 1 July 2027
Governance questions
Does CounselOS provide legal advice?
No. AI outputs are drafts and research aids only. Your lawyers and migration agents remain responsible for advice, filings, and client communication. Nothing is sent to clients until a fee earner approves it.
How should law firms use AI safely?
Use AI inside audited workflows: firm policy, human review before client delivery, protected client details, citation checking for research, and an approval inbox for drafts and billing narratives. Discourage pasting matter text into public chat tools.
Can staff use ChatGPT alongside CounselOS?
Your firm AI policy should define approved tools. CounselOS is designed so everyday drafting and research stay inside logged workflows — reducing unlogged client data in browser tabs.
What happens to client names when AI runs?
Identifying details are shielded before any outside AI service is called. Your team sees real names in CounselOS; external services receive protected placeholders mapped on our servers.
Is immigration eligibility assessed by AI?
No. Immigration assists cover drafts such as fee proposals and checklists — always reviewed by a registered migration agent or lawyer. CounselOS does not offer visa eligibility chatbots or grant-probability tools.
Walk through our AI governance posture with the team.
Immigration AI assists with drafts such as fee proposals, RFI checklists, and research memos — always reviewed by a registered migration agent or lawyer before use. CounselOS does not provide visa eligibility chatbots or grant-probability estimates.
Book a governance review