The legal AI vendor landscape, by category
Eight categories cover the legal AI market. This page describes what each category does and the diligence questions that matter before a tool touches client work. No rankings, no affiliate links, no endorsements: the same product can be safe or unsafe depending on the tier you buy, so the durable knowledge is the questions, not a leaderboard.
Legal research assistants
Generative AI layered over caselaw and statute databases: ask a question, get a synthesized answer with citations to primary law. The category includes the AI features of the major research platforms (Westlaw, Lexis) and standalone entrants.
Diligence questions: What corpus does it actually search, and is it current? Are answers grounded in retrieved documents (RAG) or generated freely? How does it present citations for verification? Embedded platform features usually inherit your existing vendor agreement; standalone tools need fresh confidentiality diligence.
Drafting and document assistants
Tools that produce first drafts: briefs, letters, clauses, memos, either general-purpose (ChatGPT, Claude, Gemini, Copilot) or legal-tuned. The same product often exists as a consumer tier and an enterprise tier with opposite data-handling terms.
Diligence questions: Does the tier you are buying train on inputs? What is retained and for how long? Can history be disabled? Treat the tier, not the brand, as the unit of trust, and verify every citation in output regardless of vendor.
Contract review and analysis
AI that reads contracts at scale: clause extraction, playbook comparison, risk flagging, due diligence sweeps. One of the most mature legal AI categories, predating the generative wave.
Diligence questions: Where do uploaded contracts live, and who can access them? Is the model trained or fine-tuned on customer documents? What accuracy benchmarks exist for your document types, and how are misses surfaced for human review?
E-discovery and TAR
Technology-assisted review and its generative successors: classifying, prioritizing, and summarizing document populations in discovery. The original "AI in law" category, with decades of caselaw around its defensibility.
Diligence questions: Is the workflow defensible and documented for court scrutiny? How are privilege calls handled and logged? Generative summarization layers add fresh confidentiality questions on top of established TAR practice.
Meeting transcription and notetakers
AI that records, transcribes, and summarizes calls and meetings. The fastest-growing tool category in firms, often adopted by staff before anyone vets it.
Diligence questions: The sharpest category for consent: recording client conversations triggers confidentiality duties and, under recent ethics guidance, informed client consent. Where do recordings live, are they used for training, and who inside the vendor can access them? Inventory these tools first; they arrive uninvited.
Intake and client-facing chatbots
Website chatbots and intake assistants that screen prospective clients, collect facts, and schedule consultations.
Diligence questions: Bot disclosure rules apply: several authorities require chatbots to identify themselves as non-lawyers. What does the bot promise, what data does it collect, where does that data go, and does its screening logic respect solicitation and UPL boundaries?
Practice management AI
AI features arriving inside existing practice management, CRM, and billing platforms: drafting assists, summaries, time-entry suggestions.
Diligence questions: The key question is whether new AI features inherit your existing agreement’s confidentiality terms or route data through new subprocessors. Vendors change these terms in product updates; re-check at each release.
Citation and brief checking
Tools that verify citations, check quotations against sources, and flag potentially fabricated authority in drafts, the direct countermeasure to hallucinated citations.
Diligence questions: What databases does verification run against? A checker is a second control, not a substitute for the signing lawyer’s own verification duty.
Frequently asked questions
Why are there no rankings or recommendations here?
Because the honest unit of evaluation is your subscription terms and use case, not a product name. The same brand can be defensible under an enterprise contract and indefensible as a free consumer account. This page maps categories and diligence questions; the evaluation is yours.
What is the single most important diligence question?
Whether the tier you are buying trains on your inputs or retains them beyond your control. That one answer separates tools that can touch client work from tools that cannot, under every ethics authority that has addressed confidentiality.
Is there an official guide to legal AI tools?
The Mississippi Bar’s "AI Tools for Lawyers: A Practical Guide" (July 2025) is the most thorough official survey, organized by category and explicitly endorsing nothing. For the binding requirements behind the diligence questions, see the Legal AI Compliance Tracker.
Useful neutral references
Educational information, not legal advice. AI terminology and tools change quickly; definitions reflect usage as of the last-updated date. For what bar associations and courts actually require of lawyers using AI, see legalaicompliance.help and consult a licensed attorney in your jurisdiction.