LawSensai AI Legal Q&A is a plain-English answer engine for the first moment of a legal problem, when you are not even sure the question you have is a legal question. You type your situation in plain language, and the engine reads it, identifies the practice area and the likely jurisdiction, and returns a short, structured answer. Every answer comes with real source citations, a safety flag that tells you how high-stakes the question is, and, when it fits, a routing card that opens the right LawSensai tool or points you toward an attorney. It returns legal information, not legal advice.
This post explains how the engine reads your situation, what comes back in an answer, how the citations work, and the guardrails that keep it honest about what an AI should and should not do alone.
How does LawSensai's AI Legal Q&A work?
AI Legal Q&A is the entry-level surface for people who have a question but do not yet know which LawSensai product they need. You type the situation in plain English or Spanish, and the engine reads three things from it.
The first is the practice area, which it selects from a fixed list that includes criminal defense, civil rights, family law, personal injury, consumer law, small claims, and estate planning, among others. The second is the jurisdiction, which it picks from the state or city you mention, or asks you for if you did not include it, because the answer often changes by state. The third is the posture, meaning the procedural stage of the matter. Someone who says they were just served reads very differently from someone who says they are only thinking about suing, and the engine treats those situations differently.
One boundary is worth stating up front. The Q&A surface answers typed questions; it does not read uploaded documents. Document reading lives inside the product workspaces, such as the Personal Injury intake or the criminal case tools, not in Q&A. Q&A is for the question itself.
The engine runs on the LawSensai agent runtime, and every answer is recorded in the hash-chain audit log, with live safety statistics published in the Trust Center. The answer you get is logged, not ephemeral.
What a structured answer contains
The engine does not return a wall of text. An answer comes back in four parts, each doing a specific job.
First, a short plain-English restatement of what your question is actually asking, so you can confirm the engine understood the situation. Second, a short plain-English answer that names the relevant rule or legal doctrine. Third, a jurisdiction note that explains how the answer changes from state to state, because that variation is often the whole story. Fourth, a safety flag, which is the single most important signal in the answer and is covered in its own section below.
The aim is for the answer to be specific enough to be useful and honest enough about its limits to be safe. It tells you what the rule is, where it varies, and how confident the engine is.
Does it cite its sources?
Yes. Every answer includes a sources list, and the sources are real. The engine cites primary sources wherever it can, meaning the actual statutes, rules of procedure, and controlling cases that govern the question. Where a primary source is not the right fit, it cites authoritative secondary sources instead.
The guardrail behind the citations matters as much as the citations themselves. If the engine cannot find a source it trusts, it says so in the answer and lowers its confidence rather than inventing one. That behavior is the point of the feature: the citations are there so you can check the answer against the underlying law, and a citation you cannot verify would defeat that purpose. An answer that admits uncertainty is doing exactly what it should.
What the safety flag means
Every answer carries a safety flag that is green, yellow, or red, and it tells you how far you should rely on the answer alone.
Green means the question is general legal information and the answer is sufficient on its own. Yellow means the question has enough specificity that you should consider having an attorney review it. Red means the question is safety-critical, and the answer is gated behind a human-attorney sign-off path. Red questions cover the highest-stakes situations, such as active criminal exposure, deportation risk, child custody emergencies, and personal injury claims that may be time-barred.
For the safety-critical situations behind a red flag, AI outputs that would go to a third party are gated. You can see the engine's draft answer in your dashboard, but the engine will not export it, copy it into a letter, or hand it to a counterparty until a network attorney has reviewed it. This is what LawSensai means by being honest about what is AI and what is human. The flag is not a disclaimer bolted on at the end; it is a routing decision that changes what the system will and will not do with the answer.
How it routes you to the right tool or attorney
When your question maps to one of the LawSensai consumer products, the answer includes a routing card that opens the right workspace, so Q&A functions as the front door to the rest of the platform. A criminal question hands off to the Criminal Defense Command Center, a family question to the Family Law Center, an injury question to the Personal Injury Recovery Center, a civil dispute to EasySuit, and a sealing or expungement question to the Record Clearing screener.
The handoff is consistent because every product shares the same audit log and the same safety-flag taxonomy. A red flag in Q&A means the same thing as a red flag in any other LawSensai product, so when the engine routes a high-stakes question into a workspace or toward an attorney, the gating travels with it. The result is that Q&A can give you a fast first answer and, when the matter needs more than an answer, move you into a place where it can be organized and, where appropriate, put in front of a real lawyer.
The guardrails, in one place
A few rules govern everything the engine does, and they are worth gathering together. The engine returns legal information, not legal advice; legal advice comes from an attorney who is licensed in your jurisdiction and has accepted you as a client. A green flag means the engine is confident the question is general, not that your case is simple, because many simple-sounding questions hide complicated facts. The engine does not read uploaded documents in Q&A; that work belongs to the product workspaces. And it cites real sources, lowering its confidence when it cannot find one it trusts rather than fabricating support.
Those guardrails are the reason the engine is useful at the very start of a problem. It will give you a clear, plain-English answer with sources you can check, it will tell you honestly how high-stakes your question is, and when the question is bigger than an answer, it will route you to the tool or the attorney that can actually carry it forward.