Ask Chatbot is LINC’s built-in assistant. You can reach it from any page on this site. It is here to help you navigate the US mental health system in your own language: finding a provider, understanding a confusing letter, drafting the right questions, or preparing for a phone call. It is not a doctor, a therapist, or a lawyer, but it can save you significant time getting ready.

If you are in crisis

Do not use Ask Chatbot in a crisis. If you are thinking of harming yourself or someone else, or you are in immediate danger:

• Call or text 988 (the Suicide and Crisis Lifeline). Interpreters are available in more than 240 languages and usually join within about a minute, and most contacts are handled with support alone, without police or an ambulance (SAMHSA).
• Call 911 for immediate danger.

If you describe a crisis in the chat, Ask Chatbot is built to recognize it and will direct you to these numbers right away. But do not wait for it. Go directly to 988 or 911.

Your privacy on LINC

LINC does not ask you to sign in or give your name, and it does not keep your conversations. When you send a message, it goes to an AI provider (Anthropic) to write the reply, and Anthropic does not use it to train its models. LINC does not collect your identity and does not share information with immigration authorities. Even so, to stay safe, do not type your immigration status, ID numbers, or full documents, and cover names and ID numbers before you share a photo of a document.

What Ask Chatbot is for

The chatbot is a navigation helper. It works best for tasks like these:

• Finding a provider in your area who speaks your language or accepts your insurance.
• Understanding a letter or form in plain language before you decide what to do.
• Drafting questions to bring to a first appointment.
• Writing a phone or email script so you know exactly what to say when you call a clinic.
• Preparing for a visit by knowing what to expect in the room.
• Asking about a LINC guide if you want a simpler explanation of something on this site.

The chatbot’s answers come from LINC’s own verified guides. This means its answers stay focused on the topics this site covers, not the open internet.

What Ask Chatbot will not do

There are things it cannot and should not do, no matter how helpful it may seem in the moment:

• Diagnose a condition or recommend a medicine.
• Provide therapy or counseling. It can help you find a therapist; it is not one.
• Give medical or legal advice. Bring those questions to a licensed professional.
• Replace a professional interpreter. For consent forms, insurance appeals, or anything you will sign, you have the right to a free, qualified human interpreter. A chatbot explanation is not a substitute for that. See your right to a free, qualified interpreter.
• Replace a person you trust. The chatbot supplements people. It does not replace them.

How to ask well

The more specific you are, the more useful the answer. A vague question gives you a general answer. A specific one gives you something you can act on.

Write in your own language. Ask Chatbot understands and responds in the language you write in. You do not need to write in English.

Say your city and state. Providers, clinics, and programs vary by location. Telling the chatbot “I live in Detroit, Michigan” helps it give you local results instead of generic ones.

Mention your insurance situation. “I have Medicaid” or “I have no insurance right now” changes which options the chatbot can point you to.

Ask one thing at a time. If you have several questions, start with the most important one and follow up after you get an answer.

Ask it to write something for you. You can say: “Write me a script I can read when I call a clinic to book an appointment.” Or: “Make me a list of questions to ask my first therapist.” The chatbot is good at this.

Paste a confusing sentence. Copy a line from a letter or a form and say: “What does this mean in simple words?” This is one of the most useful things you can do with it.

Ask for the steps, in order. You can say: “Walk me through, step by step, how to file an out-of-network claim,” or “What should I bring to my first visit?” A numbered list is easier to follow and to save.

Ask it to make it simpler. If an answer is still hard to follow, say: “Explain that again more simply,” or “Say it in shorter sentences.” You can keep asking until it makes sense to you.

The document photo feature

You can share a photo of a letter or form directly in the chat window, and the chatbot will read it and give you a plain-language explanation of what it says.

A few things to know before you use this:

• You must cover your name, date of birth, and ID numbers before taking the photo. Place a small piece of paper over those fields before you take the picture. Do not skip this step: sharing a photo with those fields visible creates a privacy risk even if LINC does not store the image.
• The image is not saved. It is sent to an AI reading service to process your question and is not stored or used to train AI models.
• It can still be wrong. Numbers, dates, addresses, and amounts are the places where AI readers make the most mistakes. Research that tested AI translation of medical instructions found it was usually accurate for common languages like Spanish, but more likely to contain at least one error in other languages, and even a small slip in a dose or a date can matter (Kong et al., 2026). So if the letter involves a deadline, a dollar amount, or a decision you must make, confirm what the chatbot says with a person and a qualified interpreter before acting on it. See making sense of insurance letters for help reading common documents.

When to switch to a person

Some decisions are too important to make based on a chatbot’s answer. This is not only caution. In hospitals, patients with limited English are more likely to be harmed by communication mistakes than English speakers (Divi et al., 2007), and federal guidance says machine translation must be checked by a qualified human when the words really matter, such as anything tied to your rights, benefits, or care (HHS Office for Civil Rights). For a simple way to sort which tasks are fine for the chatbot and which always belong to a person, see What to ask AI, and what to ask people. Use Ask Chatbot to prepare, then bring the decision to a person:

• Consent forms. You have the right to a qualified interpreter before you sign anything. Read the letter or form with the chatbot first, then confirm with a person before you sign.
• Insurance decisions. Appeals, denials, and plan choices benefit from a navigator or benefits counselor who knows your full situation.
• A diagnosis or treatment plan. Only a licensed provider can make these decisions with you.
• Anything that feels high-stakes. If you are unsure whether to trust an answer, that is a signal to bring it to a community health worker, a navigator, or your clinic’s interpreter services team. Use Ask Chatbot to get ready. Then do it with a person.

For more on using AI tools safely for health decisions, see that guide.

How to check an answer before you trust it

A good habit is to treat the chatbot’s answer as a strong first draft, not the final word. A few quick checks keep you safe:

• Ask where it came from. You can say, “Which LINC guide is this from?” and then read that guide yourself.
• Compare it with the official source. For a phone number, a deadline, or a rule, check it against the letter in your hand or the official website, like Medicaid.gov or HealthCare.gov, before you act.
• Slow down on numbers. Dates, dollar amounts, dosages, and addresses are where AI makes the most mistakes, so read those twice and confirm them with a person.
• Never sign based on a chatbot answer alone. For a consent form, an appeal, or anything you sign, use your right to a free, qualified interpreter first.
• If two sources disagree, trust the official one. The official document or your provider wins over the chatbot every time.

Doing these checks does not mean the chatbot failed. It means you are using it the smart way: to get ready quickly, then to confirm the important parts with a person.

Sources

• 988 Suicide & Crisis Lifeline. Home. https://988lifeline.org/.
• Divi, C., Koss, R. G., Schmaltz, S. P., & Loeb, J. M. (2007). Language proficiency and adverse events in US hospitals: a pilot study. International Journal for Quality in Health Care, 19(2), 60-67. https://doi.org/10.1093/intqhc/mzl069.
• Kong, M., Fernandez, A., Bains, J., Milisavljevic, A., Brooks, K. C., Shanmugam, A., Avilez, L., Li, J., Honcharov, V., Yang, A., & Khoong, E. C. (2026). Evaluation of the accuracy and safety of machine translation of patient-specific discharge instructions: a comparative analysis. BMJ Quality & Safety, 35(3), 150-158. https://doi.org/10.1136/bmjqs-2024-018384.
• Substance Abuse and Mental Health Services Administration (SAMHSA). 988 Frequently Asked Questions. https://www.samhsa.gov/mental-health/988/faqs.
• U.S. Department of Health and Human Services, Office for Civil Rights. Section 1557: Ensuring Meaningful Access for Individuals with Limited English Proficiency. https://www.hhs.gov/civil-rights/for-individuals/section-1557/fs-limited-english-proficiency/index.html.

LINC is a research prototype, not a medical or legal service. Translation can be wrong; you have the right to a free, qualified interpreter. For anything high-stakes, do it with a community health worker or someone you trust. In a crisis, call or text 988, or 911 for immediate danger.