Two people can ask the same question and need opposite answers. "Can I be evicted for late rent?" from someone who paid four days late is a question about cure periods. The same sentence from someone holding a summons with a hearing date is a question about deadlines, and answering it with general information about cure periods could cost that person the case.
That is why the LawSensai AI Legal Q&A engine does not treat your question as a search string. Before it writes anything, it reads three things out of your situation: the practice area, the jurisdiction, and the posture. This post explains what each one does and why the third is the one that most changes the answer.
What is the engine reading first?
Practice area is picked from a fixed list rather than inferred loosely. Criminal defense, civil rights, family law, personal injury, consumer law, small claims, estate planning, and several others. A fixed taxonomy sounds like a limitation and is actually the opposite: it is what lets a question map to a specific workspace instead of to a generic answer. When the engine classifies your situation as consumer law with a lawsuit in hand, there is a defined product on the other side of that classification.
Classification also handles the common case where a situation spans two areas. A layoff involving a signed severance agreement is partly employment and partly contract. That kind of question is exactly what Sensai Concierge exists to route, because a category menu forces a choice the user should not have to make and often gets wrong.
Jurisdiction is picked from the state or city you mention. If you did not mention one, the engine asks rather than defaulting. That refusal to guess matters more in American law than in almost any other domain, because the substantive rule, the filing deadline, and the available remedy are frequently all set at the state level. Every answer carries a jurisdiction note explaining how the analysis changes across states, and answers that depend on an unconfirmed state are answers built on sand.
Why does posture change the answer so much?
Posture is the procedural stage you are actually at. Thinking about a claim. Preparing to send a demand. Already sued. Already defaulted. Post judgment. Appeal window running.
Posture is the variable that general legal content on the internet almost never captures, because an article is written once for everyone. But posture determines three things at once.
It determines the clock. Someone considering a personal injury claim is working against a statute of limitations measured in years. Someone who has been served with a complaint is working against an answer deadline measured in days, and in most debt collection cases the failure that ends the case is a missed date rather than a weak defense.
It determines what is still available. Defenses that are decisive before judgment are frequently unavailable after it. Negotiating leverage that exists before filing shrinks once the other side has spent money on a lawyer. A demand letter is a strong opening move and a weak response to a summons.
It determines the right next step. The same underlying dispute routes to a demand letter generator at one posture, to a defense triage and an answer draft at another, and to a post judgment path at a third.
Reading posture is what lets the engine answer the question you actually have instead of the topic you happened to name.
What does the answer look like once those three are set?
Answers come back in four parts, and the structure is fixed on purpose.
A short plain English restatement of what you are actually asking, so you can catch a misread immediately rather than after acting on it.
A short answer that names the governing rule or doctrine, rather than gesturing at it. Naming the rule is what makes the answer checkable and what makes it useful to bring to an attorney.
A jurisdiction note explaining how the answer shifts by state, which is the honest form of an answer to most American legal questions.
A safety flag, green, yellow, or red. Green means general legal information is likely sufficient for what you asked. Yellow means the question is specific enough that attorney review is worth considering. Red means the question is safety critical, and the answer is gated behind a human attorney sign off path. If you read one element of the response before acting, read the flag.
What happens when the engine does not have enough facts?
It says so. The engine is built to be explicit about uncertainty rather than to fill the gap with a confident sentence, and it does not predict outcomes. It will not tell you what your case is worth or whether you will win, because those are predictions that no responsible general information tool can make and that a licensed attorney makes only with the full file in front of them.
In practice, missing facts usually take one of three forms: the state was never stated, the posture is ambiguous, or a date that controls a deadline is unknown. In each case the engine asks rather than assumes, and the follow up question is usually short.
How does the answer connect to the rest of the product?
When the classified situation maps to a product, the answer hands you a routing card that opens the workspace built for it. EasySuit for a lawsuit in either direction, the Family Law Center for a matter that will live for months, the Personal Injury Recovery Center for an accident, Credit Defense for a report or a collector, the IP Center, Business Formation, the Licenses and Permits Finder, and so on.
The routing card is the joint between the answer layer and the work layer. It exists because the most common failure mode of legal information is not a wrong answer, it is a correct answer that leaves the reader with no idea what to do next.
You are never forced through it. Many people get what they need from the answer alone and stop there, which is a legitimate outcome and by design.
What this does not do
Classification is not representation. Reading your practice area, jurisdiction, and posture accurately produces better general legal information, not legal advice, and LawSensai is not a law firm. The AI disclosure on every answer is there because the distinction is real.
The engine also cannot verify facts you supply. If the date of service you enter is wrong, the deadline computed from it will be wrong in the same direction. Documents and dates are worth double checking against the paper in your hand before relying on any calculation.
And a green flag is not a clearance. It means the question, as asked, is one that general legal information addresses well. If your situation has features you did not mention, the flag was set on an incomplete picture.
The practical takeaway
If you want a better answer from LawSensai, or from any legal research for that matter, give it the three inputs it is looking for. Name your state. Name where you are in the process, including any date on any paper you have received. Describe the situation rather than the legal label you think applies, because the label is the part the engine is good at and the facts are the part only you have.
Most bad legal answers on the internet are not wrong. They are answers to a different posture.