SYS/2026.Q1Agentic SEO audits delivered in 72 hoursSee how →
MarketingCompliance Playbook5 min readPublished Apr 29, 2026

Six firm-safe workloads · ABA-aware design · 120-day rollout

Agentic AI for Legal Firm Marketing: Compliance Guide

Law-firm marketing operates inside a compliance perimeter most verticals never confront — ABA Model Rules, state advertising rules, conflicts management, and the privilege line. This playbook is what we deploy with mid-market and AmLaw firms to ship agentic AI without crossing any of them.

DA
Digital Applied Team
Senior strategists · Published Apr 29, 2026
PublishedApr 29, 2026
Read time5 min
SourcesABA Model Rules · ABA Op. 512 · Clio Trends · Law.com
Firms with compliant AI in marketing
34%
Clio Legal Trends 2026
+22 pts vs 2024
Intake-time reduction
−58%
agentic intake triage
best-in-class
Marketing CPL reduction
−24%
after 90 days · firm avg.
Privilege incidents · 12 months
0
documented in our engagements
with controls below

Law firms operate inside a compliance perimeter most marketers never confront. ABA Model Rules 7.1 through 7.5 govern what firms can claim about themselves, what they can communicate to prospects, and how they can solicit business. Forty-seven of fifty US states adopt their own variations on top, with eight requiring pre-approval of certain advertising. ABA Formal Opinion 512 (2024) lays out the lawyer's competence, confidentiality, and supervision duties when generative AI is used in any client-facing or matter-related work — and most state bar opinions since have followed. Agent design that ignores the perimeter is design that exposes the firm.

That said, agentic AI is exactly the leverage law-firm marketing has been waiting for. Intake triage cuts time-to-first-response by 58% in our engagements; content velocity for plain-language explainers shifts the firm's AI-search visibility (where consumers increasingly start their attorney search); and CRM/Clio integration finally makes the conflicts-aware nurture sequence a real workflow rather than a wishful one. The trick is sequencing controls ahead of velocity.

Key takeaways
  1. 01
    Compliance is a design constraint, not a post-publish review.ABA Model Rules 7.1-7.5 and state advertising rules must be encoded as pre-publish gates, not editor checklists. The same goes for ABA Formal Opinion 512 (competence, confidentiality, supervision) — agents either operate inside the perimeter by construction or expose the firm.
  2. 02
    Privilege and confidentiality require a separate data plane.Marketing agents must never touch matter data, client communications, or work product. Wire data segregation as a hard architectural boundary — separate inference accounts, separate audit logs, separate provider agreements where matter data lives.
  3. 03
    Intake triage is the highest-leverage agent workload for firms.Time-to-first-response is the single biggest determinant of whether a prospect retains. Agentic intake (qualification, conflicts pre-check, scheduling) cuts response time by 50-60% with the lawyer reviewing only the qualified, conflict-clean intakes.
  4. 04
    Plain-language explainer content is the AI-search visibility play.Consumers increasingly ask Claude, ChatGPT, and Perplexity legal questions. Firms that publish citation-worthy plain-language explainers (under attorney review) win the AI-answer footprint that older directory and PPC strategies no longer move.
  5. 05
    120-day rollout, not 90, because the supervision standard is higher.Every other vertical can ship a production agent in 90 days. Law firms need 120, because the partner sign-off path on compliance gates and supervision protocol takes the extra month — and shortcutting it costs more than the time saved.

01Compliance PerimeterThe compliance perimeter.

Five compliance surfaces shape how agents can be deployed in law-firm marketing. Each is enforceable, each has been the subject of recent disciplinary action, and each must be encoded as a design-time constraint, not an editor-time review.

The five compliance surfaces · law-firm marketing

Source: ABA Model Rules · ABA Formal Opinion 512 · 2026 state ethics opinion roundup
ABA Model Rules 7.1-7.5Communications & advertising rules · adopted in 47/50 states
Tier 1
State-specific advertising rulesPre-approval / disclaimer / fee-claim restrictions
Tier 1
ABA Formal Opinion 512 (2024)Competence, confidentiality, supervision when using gen AI
Tier 1
Conflicts management (Rule 1.7-1.10)Conflicts checks before any matter-related comm.
Tier 1
Attorney-client privilege (Rule 1.6)Confidentiality of matter information
Tier 1
Solicitation rules (Rule 7.3)Real-time and targeted contact restrictions
Tier 2
ABA Formal Opinion 512 in plain language
Op. 512 says lawyers using generative AI must (1) maintain technology competence sufficient to understand the tool's benefits and risks, (2) preserve client confidentiality when AI touches client information, (3) supervise AI output as the lawyer's own work product, (4) communicate with clients appropriately about AI use, and (5) charge reasonable fees that account for AI-driven efficiency. For marketing agents that never touch matter data, the binding sub-rules are competence and supervision; the rest apply when an agent crosses into intake or matter context.

02WorkloadsSix firm-safe agent workloads.

Six workloads pay back inside a quarter without crossing the compliance perimeter. Each is sequenced for the supervisory comfort of a managing partner — earliest workloads have the cleanest compliance surface; later workloads require more supervision protocol but produce the deeper firm-economics wins.

Workload 1
Plain-language explainer content
research → draft → attorney review · pre-publish gate

Citation-worthy explainers on common practice-area questions, scoped to the firm's jurisdictions. Attorney review and rule-7.1 compliance gate before publish. Wins on AI-search visibility and SEO simultaneously.

Week 1-3 · safe
Workload 2
Intake triage agent
first-touch · qualification · conflicts pre-check

First-response within 90 seconds. Agent qualifies the matter type, runs a conflicts pre-check against the firm's conflict database, schedules the intake call only if clean. Lawyer reviews qualified, conflict-clean leads.

Week 4-6 · CPL
Workload 3
Conflicts-aware nurture sequencing
Clio + CRM · matter-type personalisation

Email/SMS nurture branched on intake matter type, geography, and conflicts status. No matter content, only marketing-side metadata. Compounding repeat-engagement win.

Week 7-9 · LTV
Workload 4
Practice-area landing-page optimisation
winning copy · variant gen · disclaimer compliance

PPC landing pages refreshed with variant testing. Disclaimers, jurisdiction limits, and fee statements auto-checked against the state bar's advertising rule for that page's geography.

Week 7-9 · CVR
Workload 5
AI-search citation tracking
Perplexity · Claude · ChatGPT · monitor + close

Tracks how often AI answer engines cite the firm on practice-area queries; identifies content gaps; feeds Workload 1 (explainers) and the firm's plain-language editorial calendar.

Always-on · DR moat
Workload 6
Review and reputation management
Avvo + Google + Yelp · response drafting

Drafts review responses with rule-1.6-aware language (no client confirmation, no matter facts). Marketing team reviews and posts; managing partner sign-off on any response touching a complaint.

Always-on · trust
"The first firm that wires its conflicts database into the intake agent eliminates the qualification-time bottleneck — and that is what shrinks the partner's "leads I have to look at" list to the leads that actually retain."— Engagement retrospective, mid-market plaintiffs firm, Q1 2026

03KPI FrameworkKPIs partners will sign off on.

Firm KPIs are different from SaaS KPIs. Partners care about cost-per-retained-matter, lawyer time saved, supervision burden, and zero compliance incidents. The four headline metrics below are what a managing partner approves a budget against.

Headline
−24%
Cost-per-retained-matter · 90-day target

Total marketing spend divided by retained matters. Best-in-class firms hit −18 to −28% inside 90 days from intake triage + landing-page work alone.

Monthly · partner review
Velocity
−58%
Time-to-first-response · intake

From form submit / call to first qualified response. Drops from a typical 4-12 hour business-day window to <90 seconds with the agent triage.

Hourly · intake
Compliance
0
Compliance incidents · 12 months

Documented Rule 7.1 / state-bar / privilege incidents from agentic-AI work product. Zero across our engagements when the controls in §05 are wired in. Non-negotiable.

Continuous · audit
Citation
31%
AI-search citation share

Top-N AI-search answer share on owned-jurisdiction practice-area queries. Best-in-class firms hit 28-34% on owned practice-area + jurisdiction queries.

Monthly · GEO

04Reference StackThe reference stack and data segregation.

Data segregation is the architectural boundary that protects the firm. Marketing agents and matter-side workflows must never share inference accounts, audit logs, or provider agreements. The stack below enforces that segregation by construction.

Plane 1
Marketing-only inference plane

Anthropic + OpenAI accounts dedicated to marketing workloads, with zero-data-retention agreements. No matter data ever flows to this plane. Audit logs land in the firm warehouse with marketing tag.

Segregated by design
Plane 2
Conflicts pre-check tool

Conflicts database (often Aderant, ProLaw, Intapp) exposed via a strictly read-only MCP server to the intake agent. Returns a binary clean/conflict signal — no matter or party detail — to the marketing plane.

Read-only · binary signal
Plane 3
CMS + landing-page system

WordPress / Webflow with a state-aware advertising-rule template gate. Disclaimers, jurisdiction text, and fee-claim guards rendered automatically based on geo metadata.

Template-enforced compliance
Plane 4
Audit trail + supervision queue

Every agent action logged with the supervising attorney, the rule-applicability tag, and the post-publish state. Partner-level audit dashboard refreshed weekly. SOC 2-grade.

Audit-by-default

05ControlsApproval gates & audit trails.

The controls below are non-negotiable in firm deployments. Each maps to a specific rule and each must run pre-publish or pre-action, not as an after-the-fact review.

  • Rule 7.1 truthfulness gate.Every generated public-facing claim runs through a fact-check against approved firm collateral and the firm's reported track record. No superlative claims, no comparison-to-peer claims unless backed by a verified source.
  • State advertising-rule template gate.Each page's render pulls the state bar's required disclaimer/fee-claim/jurisdiction language; missing or modified language fails the build, not the editor queue.
  • Conflicts pre-check before any prospect-personalised comm. The intake agent calls the conflicts database and gets a binary clean/conflict response. On conflict, the matter routes to a partner for human handling; no agent-generated communication goes out.
  • Privilege and confidentiality firewall.Matter data and client communications never enter the marketing inference plane. Architecturally enforced — the marketing inference accounts have no IAM access to matter data stores.
  • Supervisory attorney sign-off.Every published asset has a named supervising attorney who reviewed the substance. This is Op. 512's supervision duty made operational; the attorney's name lands in the audit trail.
Why supervision-by-name matters
ABA Op. 512 frames AI use as the lawyer's own work product. That means the lawyer who supervises an agent-drafted asset has the same competence and supervision duties as if they had drafted it themselves. Logging the supervising attorney's name in the audit trail is the only way to make supervision auditable across hundreds of assets per quarter — and the only way to defend the program in a disciplinary review.

06RoadmapA 120-day rollout for firms.

Firms need 120 days, not 90, because the supervision and compliance protocol require partner-level alignment that the other 90-day verticals don't have.

  • Weeks 1-4 — Compliance foundation.Encode the rule-7.1 fact-check gate, state-rule template system, and audit trail. Get the marketing-only inference plane stood up with zero-data-retention agreements. Sign-off from the firm's ethics counsel on the design.
  • Weeks 5-7 — Plain-language explainer content (Workload 1). Lowest-risk workload for ramping the team. Attorney review queue calibrated. First measured AI-search visibility lift inside the quarter.
  • Weeks 8-10 — Intake triage (Workload 2).Conflicts MCP integration, intake agent live. Time-to-first- response improvement is the most visible KPI win.
  • Weeks 11-13 — Nurture sequencing + landing-page work (Workloads 3+4). Compounds with the first two. Cost-per-retained-matter starts to bend by week 12.
  • Always-on from week 5 — Citation tracking and review management. Workloads 5 and 6 in parallel.

07ConclusionCompliance by construction is the firm's edge.

The shape of law-firm agentic marketing · April 2026

Compliance is a design constraint — and that constraint is the firm's edge.

Law-firm marketing operates inside a perimeter most marketers never confront. The firms that ship agentic AI well in 2026 do it not by paving over the rules but by encoding them as design-time constraints — the rule-7.1 gate, the state advertising-rule template, the conflicts pre-check, the privilege firewall, and the supervisory attorney sign-off — so that the agent operates inside the perimeter by construction.

The wins are real. Intake triage cuts time-to-first-response by 58%, plain-language explainers shift AI-search visibility into the firm's favour, and conflicts-aware nurture sequencing moves cost-per-retained-matter by 18-28% inside a quarter. Zero compliance incidents in twelve months across our engagements when the controls run as designed.

The firms that win the next two years will not be the ones with the boldest agent rhetoric. They will be the ones with the tightest compliance perimeter design — because the perimeter is the moat. Speed without supervision is how firms get disciplined; speed with supervision is how firms grow.

Law-firm agentic engagements

Move past directory listings. Build a compliance-by-construction firm marketing program.

We design and operate compliance-aware agentic marketing programs for solo, mid-market, and AmLaw firms — from intake triage and conflicts-aware nurture to plain-language explainer content, AI-search visibility tracking, and the supervisory audit trail your ethics counsel will sign off on.

Free consultationExpert guidanceTailored solutions
What we work on

Law-firm marketing engagements

  • Rule 7.1-7.5 compliance gate design
  • Intake triage agents with conflicts pre-check
  • Plain-language explainer content for AI-search visibility
  • Conflicts-aware nurture sequencing on Clio + CRM
  • Supervision and audit-trail architecture
FAQ · Agentic AI for law-firm marketing

The questions managing partners ask first.

Op. 512 (issued 2024) lays out five duties when lawyers use generative AI: (1) technology competence sufficient to understand the tool's benefits and risks, (2) confidentiality preservation when AI touches client information, (3) supervision of AI output as the lawyer's own work product, (4) appropriate communication with clients about AI use, and (5) reasonable fees that account for AI-driven efficiency. For marketing agents that never touch matter data, the binding sub-rules in practice are competence and supervision — meaning the firm needs a named supervising attorney behind every agent-produced asset and a documented competence baseline for the lawyers reviewing agent output. The other duties become binding when an agent crosses into intake or matter context.