T52 AI adoption readiness assessment
AI adoption readiness assessment
Where Ashgrove Legal Partners is ready to adopt AI safely today, where the firm is not yet ready, and where to start. Prepared by Northbeam AI Advisory following a two-week assessment: 11 stakeholder interviews, a review of Ashgrove's document systems and a scan of current tool usage across four practice groups.
How to read this assessment
Ratings below reflect Ashgrove's readiness on each dimension, not a judgment of the firm's competence. A low rating on Data and Systems means the underlying systems need work before AI use is safe there, not that staff are behind. All ratings are Northbeam's professional judgment from the interviews and system review, not an automated score.
Sample template: the firm, people, clients and figures are fictional and shown for illustration only.
Readiness by dimension
Readiness scores as progress
Governance is the firm's weakest dimension: there is currently no written policy on what client information may enter any external tool, which blocks safe use of most use cases below regardless of enthusiasm elsewhere.
Evidence
- Ray Kowalski has publicly committed to exploring AI at two firm meetings since June 2026.
- Three of four practice group leads support a pilot; the real estate group lead is skeptical, citing client confidentiality.
- No budget has yet been allocated beyond this assessment.
- Document management is split across a legacy DMS and a shared drive with inconsistent folder structure.
- No client matter data is tagged by confidentiality level, so automated access controls are not currently possible.
- Billing and time entry are in a separate system with no integration to the DMS.
- 6 of 11 interviewed attorneys already use a general AI assistant personally for drafting, unofficially.
- Paralegals report the most interest in AI for first-pass document review.
- Two senior partners expressed concern about AI-generated errors in client-facing documents.
- No written AI use policy exists; personal use of AI tools with client data is currently unmonitored.
- No process for reviewing AI-assisted work product before it reaches a client.
- Malpractice insurer has not yet been consulted on AI use, per the managing partner.
Use case shortlist
| Use case | Practice group | Readiness | Data sensitivity | Recommended start |
|---|---|---|---|---|
| First-pass review of standard commercial leases | Real estate | Medium, client-identifying | After governance policy is signed | |
| Internal drafting of non-client memos and summaries | All groups | Low, internal only | Immediate pilot candidate | |
| Client-facing contract drafting | Commercial | High, client-specific terms | Hold until human review workflow is defined | |
| Time entry narrative drafting from calendar and notes | All groups | Low, internal only | Immediate pilot candidate | |
| Litigation document review at scale | Litigation | High, privileged material | Hold until data tagging exists |
Guardrail required: no policy on client data in external tools
Six attorneys already use general AI tools informally with no policy governing what client information may be entered. We recommend the firm publish an interim policy within two weeks that names which tools are approved and what data categories are prohibited, before any formal pilot expands usage. Owner: Ray Kowalski and the firm's malpractice insurer contact. If not addressed before the pilot begins, informal use will likely continue unmanaged in parallel with any sanctioned pilot.
Guardrail required: no human review checkpoint defined
No use case above should reach a client without a named attorney sign-off step. Ashgrove does not currently have a documented review checkpoint for any AI-assisted work product. Owner: Practice group leads, one checkpoint per use case, before pilot launch.
30, 60, 90 day plan
Days 1 to 30. Publish the interim AI use policy. Launch internal drafting and time entry narrative pilots with the paralegal group, since both are low sensitivity and ready today.
Days 31 to 60. Begin tagging real estate matter documents by confidentiality level. Start the guarded lease review pilot with two attorneys and a defined human sign-off step before any output reaches a client.
Days 61 to 90. Review pilot results against the readiness ratings above, decide which use cases move to firm-wide rollout and which need more governance work before scaling.
How to use this template
Duplicate this page for each client AI readiness assessment. Replace every rating and evidence item from the actual interviews and system review; if a dimension has not been assessed, write Needs review instead of assigning a rating.
| Variable | Meaning | Required |
|---|---|---|
{{client.name}} | Client the readiness assessment covers | Required |
{{assessment.window}} | Dates the interviews and system review covered | Required |
{{dimension.ratings}} | Rating per readiness dimension, 1 to 5, with evidence | Required |
{{usecase.shortlist}} | Candidate use cases with readiness, data sensitivity and recommended start | Required |
{{guardrail.required}} | Guardrails that must exist before any pilot expands, with owner | Required |
{{plan.phase_actions}} | Actions for each 30-day window | Optional |
Never assign a numeric readiness rating without interview or system evidence behind it; mark an unassessed dimension as Needs review instead.