Smokeball AI Review for the Small Law Firm
Smokeball is a practice management platform built primarily for small and solo firms in the US, UK, and Australia. Over the past two years it has layered a set of AI features — most visibly AI Field Population, Archie (an AI assistant), and matter/document summarization — on top of what has long been a well-regarded document automation and passive time-tracking suite. This review looks at Smokeball on its own terms, then puts its data flows under the confidentiality lens that practicing lawyers actually care about.
Disclosure: Basil, which publishes this review library, competes with Smokeball in the meeting-notetaker category. Basil is a fully on-device notetaker; Smokeball is a full practice management suite with some AI functionality bolted in. Those are different products, and this review treats Smokeball as what it is: a PMS. If you want to see how our approach differs on the narrow question of meeting capture, see Basil for Law.
What Smokeball Actually Does
Smokeball's core is a Windows-first (with growing web and mobile) practice management platform: matter management, contacts, calendaring, document assembly from a large template library, automatic time capture, trust accounting, billing, and integrations with Microsoft Word, Outlook, and QuickBooks. Its historic differentiator has been passive time tracking — the client watches which documents and emails you touch inside a matter and reconstructs billable time — plus a very deep Word/Outlook integration. See the features page for the current scope.
On top of that, Smokeball has added AI features under an umbrella they market as Smokeball AI. The pieces most relevant to daily practice:
- AI Field Population — extracts data from client intake documents or uploaded PDFs and drops it into matter fields, cutting a lot of retyping.
- Archie — a chat-style assistant that can summarize a matter, draft emails, and answer questions grounded in the matter file.
- Document and email summarization — one-click summaries of long documents and threads.
- Communicate — a client portal (not AI, but part of the value stack most small firms will use alongside the AI features).
None of this is revolutionary in isolation, but the integration story is genuinely strong: because Smokeball already sits between Word, Outlook, and the matter file, its AI features have real context to work with rather than needing you to paste content back and forth.
The Confidentiality Lens
This is the section that matters most. Smokeball is a cloud practice management platform. Client data — matter files, documents, emails synced from Outlook, contacts, billing entries — is stored on Smokeball's cloud infrastructure. The desktop app is a client on top of that cloud store, not a local vault.
Key architectural facts, from Smokeball's own materials:
- Hosting. Smokeball states it hosts customer data on Microsoft Azure, with encryption in transit and at rest and SOC 2 controls. Data location depends on the region you sign up in.
- AI processing. The AI features rely on third-party foundation models (Smokeball has publicly discussed using OpenAI and Azure OpenAI infrastructure). That means when you invoke Archie or a summarization feature on a matter document, that content is sent from Smokeball's environment to a model provider for inference. It is not processed locally on your machine.
- Training use. Smokeball's terms and AI documentation indicate customer content is not used to train third-party foundation models. Confirm the current version at signing; contract language changes.
- Retention. Data lives in your Smokeball tenant until you delete it or terminate the account, subject to the deletion timelines in the terms. There is no automatic short-term purge of AI inputs the way some standalone tools offer.
- DPA / BAA. Smokeball offers a Data Processing Addendum. A BAA for HIPAA-covered engagements is not something the vendor broadly markets; firms with health-related matters should ask before assuming coverage.
What a subpoena served on Smokeball could reach. Because Smokeball is the custodian of your matter files in its cloud, a subpoena or court order served on the vendor could, in principle, reach the stored documents, emails synced into matters, contacts, and time entries — the same exposure profile as any cloud PMS (Clio, MyCase, PracticePanther). Whether privilege and work-product doctrine defeat production is a separate legal question, but the data is there to be asked for. This is not unique to Smokeball; it is the price of admission for cloud practice management. Compare against a local-first tool where the vendor simply does not hold the content.
The narrower AI question — whether prompts and outputs are logged in a way that could be produced — is worth asking your rep in writing. The Brewer v. Otter.ai litigation is a reminder that AI-adjacent transcripts and logs are discoverable artifacts, not ephemeral.
Ethics-Opinion Fit (ABA 512)
ABA Formal Opinion 512 (July 2024) sets out the current baseline for generative AI use by lawyers: competence in the tool, informed client consent where confidential information is disclosed to a third-party model, supervision of outputs, and reasonable due diligence on the vendor. NYC Bar Formal Opinion 2025-6 (December 2025) pushed the client-consent conversation further, emphasizing that meaningful consent requires clients to understand where their data is going.
How Smokeball maps to that:
- Competence. The tool is well-documented and the AI features are narrow enough (field extraction, summarization) that competent supervision is achievable. Archie's outputs still need lawyer review — treat every draft as a first-year associate's work product.
- Confidentiality. Because AI processing happens off-device on third-party model infrastructure, engagement letters should disclose that cloud practice management and AI subprocessors are in use. Model Rule 1.6(c) requires reasonable efforts; a SOC 2 vendor with contractual no-training clauses is defensible, but it is not a substitute for client-facing disclosure.
- Supervision. AI Field Population is the highest-risk feature to leave unchecked, because a wrong party name or date silently populated into a matter can propagate into documents. Build a review step.
- Vendor diligence. Ask for the current subprocessor list, DPA, and a written statement on training use before rollout.
None of this rules Smokeball out. It is the same diligence any responsible firm should be doing with Clio Duo, MyCase IQ, or any other cloud PMS with AI features. Compare with the reasoning in West Technology Group v. Sundstrom and the more recent US v. Heppner (S.D.N.Y. Feb 2026), where Judge Rakoff held a litigant's chats with a public AI platform were not privileged. The lesson: architecture matters, and a lawyer-tenant SaaS with a DPA is materially different from a consumer chatbot, but both need to be understood before you type client facts into them.
Where Smokeball Is Genuinely Strong
Credit where it is due. Smokeball's document automation library and Word integration are among the best in the small-firm segment. Passive time capture is a real revenue lift for firms that historically undercount billable minutes. The AI Field Population feature is a legitimately useful piece of workflow AI — the kind that removes drudgery without pretending to practice law. Archie's matter-scoped answers are more grounded than a general-purpose chatbot precisely because they draw from the file rather than the open internet.
For a firm that lives in Word and Outlook and wants a single tool for matters, documents, time, and billing, Smokeball is a serious contender. It is not trying to be everything; it is trying to be the operating system of a small firm, and it is largely succeeding.
Where It Falls Short
The Windows-first legacy still shows. Mac users have historically been second-class citizens, though the web app has narrowed the gap. Reporting is functional but not analytical. Pricing is quote-based rather than transparent, which makes comparison shopping harder. And the AI features, while useful, are not differentiated on confidentiality architecture — they use the same cloud-model pattern as competitors, so if you were hoping for on-device or zero-retention AI as a competitive edge here, that is not what Smokeball is offering.
For meeting capture specifically, Smokeball's ecosystem does not include an on-device notetaker. Firms concerned about client conversations being sent to third-party transcription services will need to pair it with something else; Basil is one option, but so is disciplined use of a local recorder plus manual notes.
Pricing and Who It's For
Smokeball uses tiered plans (Bill, Boost, Grow, Prosper+) with per-user pricing that is not publicly listed in full — you get a quote. The pricing page lays out the tier structure and what is included. AI features are bundled into the higher tiers rather than sold separately.
Who it fits:
- Solo and small firms (1–20 lawyers) in litigation, family, estate, real estate, or immigration who want deep document automation and billing in one place.
- Firms that live in Word and Outlook and want passive time capture to stop leaving money on the table.
- Firms comfortable with cloud PMS and willing to do subprocessor diligence.
Who it does not fit:
- Firms with a hard local-only or air-gapped confidentiality posture.
- Very large firms with enterprise DMS needs (iManage, NetDocuments territory).
- Firms that want a la carte AI tools rather than a bundled suite.
Verdict
Smokeball is a mature, well-integrated small-firm PMS with a sensible and improving AI layer. The AI features reduce drudgery rather than overreach, which is the right posture. The confidentiality profile is standard cloud-SaaS: adequate for most firms with proper client disclosure and vendor diligence, but not a fit for firms that need client data to stay off vendor servers entirely.
| Pros | Cons |
|---|---|
| Excellent document automation and Word/Outlook integration | Cloud-only architecture; vendor holds all matter data |
| Passive time capture recovers real billable minutes | AI processing routed to third-party foundation models |
| AI Field Population and Archie are practical, matter-scoped | Pricing is quote-based, not transparent |
| SOC 2 controls, DPA available, contractual no-training terms | BAA availability limited; confirm before health-adjacent work |
| Strong fit for small firms in Word/Outlook workflows | Windows-first heritage still shows on Mac |
Bottom line: if you are a small firm looking for one system to run matters, documents, time, and billing, Smokeball earns a place on the shortlist. Do the vendor diligence, update your engagement letter, and treat AI outputs as drafts. For meeting capture with a stricter confidentiality architecture, evaluate a dedicated on-device tool like Basil alongside it.
This review is for information only and is not legal advice.
Frequently asked questions
Does Smokeball train AI models on my client data?
Smokeball's terms and AI documentation state that customer content is not used to train third-party foundation models. Because contract language changes, confirm the current version with your account team at signing and get the no-training commitment in writing.
Where is Smokeball data stored?
Smokeball hosts customer data on Microsoft Azure, with the specific region depending on where you signed up (US, UK, or Australia). Encryption in transit and at rest and SOC 2 controls are in place per Smokeball's security page.
Can a subpoena served on Smokeball reach my client files?
Yes, in principle. Because Smokeball is the cloud custodian of matter files, documents, and synced emails, a subpoena or court order to the vendor could reach that stored content. Privilege and work-product doctrine may still block production, but the data exists on vendor infrastructure. This is the same exposure profile as any cloud PMS.
Does Smokeball satisfy ABA Formal Opinion 512?
ABA 512 is a lawyer-conduct standard, not a vendor certification, so no tool 'satisfies' it on its own. Smokeball provides the infrastructure (SOC 2, DPA, contractual no-training terms) that supports a defensible position, but the lawyer must still handle client consent, supervise outputs, and conduct vendor diligence.
Is Smokeball a good fit for solo lawyers?
Yes, particularly for solos who bill hourly, live in Word and Outlook, and want document automation, passive time capture, and billing in one system. Solos with a strict local-only confidentiality posture, or who need enterprise DMS features, should look elsewhere.
How does Smokeball compare to Basil for meeting notes?
They are different categories. Smokeball is a practice management suite whose AI focuses on matter documents, field extraction, and summarization. Basil is a dedicated on-device meeting notetaker that processes audio locally on the lawyer's machine. Firms often pair a PMS with a separate notetaker rather than choose between them.
Meeting notes with no server to subpoena
Basil transcribes and summarizes entirely on-device — privilege-safe by architecture. See Basil for Law →
This review is for information only and is not legal advice.