AI disclosure

How we use AI

When you call our main number or use the chat box on this website, you are talking to a computer program, not a person. This page explains what it is, what it can and cannot do, and how to reach a member of our staff.

You are talking to an AI, not a person

Our main phone line is answered by an AI assistant. Its name is Eva. The chat box on this website is answered by the same kind of automated assistant. Neither one is a human being.

Eva is not a doctor, a nurse, or a psychologist, and she is not Dr. Barnoski. She says she is automated at the start of every call, and the chat says so too. If you ever want to check, just ask — she will tell you straight away.

You can ask for a person at any time. Call our office on (631) 366-3369 during office hours, Monday – Friday, 9:00 a.m. – 5:00 p.m., and say you would like to speak to a member of staff.

Scope

What the AI can and cannot do

The assistant is a receptionist, not a clinician. It is deliberately limited, and it is meant to hand you to a person the moment a question goes past those limits.

It can

  • Answer general questions about the practice — what we do, who we see, and what we do not offer
  • Tell you our office hours, our address, and how to get here and park
  • Explain which insurance plans we take part in, and what a referral needs to say
  • Take a message with your name, your phone number, and why you are calling, and pass it to the right member of staff
  • Send you a link to our consultation request form, or help you start one
  • With your permission, take the details from your insurance card and run an electronic check on your coverage

A coverage check tells you what your insurance company reports. It is not a promise of payment, and our staff confirm your benefits before your visit.

It cannot

  • Give medical advice, or tell you what is wrong with you
  • Diagnose anything, or say whether you have a condition
  • Explain, interpret, or comment on test results, scores, or a report
  • Book, move, or cancel an appointment — only our staff do that, and they will call you
  • Help in an emergency, or provide crisis or mental-health support
  • Give legal advice, or tell you what your bill will finally be
  • Confirm details about an existing patient's care, or read anything back to you from a record

If you ask about any of these, the assistant will say it cannot help and take a message for a person instead. That is the correct behavior, not a fault.

In an emergency, do not use the AI

If you are experiencing a medical emergency, call 911 or go to the nearest emergency department. This practice does not provide emergency or crisis services.

The assistant cannot help you in an emergency, and no one is listening to the chat box around the clock. If someone is in danger, hang up and call 911, or go straight to the nearest emergency department.

If you are thinking about harming yourself or taking your own life, help is available at any hour, free of charge: call or text 988, the 988 Suicide & Crisis Lifeline, or call 911. If someone tells our assistant something of this kind, it is set up to stop what it is doing and give the emergency instruction above immediately, rather than carrying on with a message or a form.

Our AI policy in full

Reaching a person

You never have to deal with the AI. Ask for a person, say "staff", or simply call during office hours and tell us you would rather speak to someone. Our office number is (631) 366-3369, and we are open Monday – Friday, 9:00 a.m. – 5:00 p.m..

If you call when the office is closed, or when everyone is with a patient, the assistant will take a message and pass it to the right member of staff, who will call you back. You can also tell us that you would prefer we never use the automated assistant with you, and we will note that.

Relatives, carers, and referring physicians can do the same. Nobody is required to talk to a machine to reach this practice.

Calls are recorded

Calls to our main line are recorded. The assistant says so at the very beginning of the call, before anything else, so you can decide whether to carry on or to call back and ask for a person.

We record for three reasons: so that the message reaching our staff is accurate and nothing you said is lost; so that we can check the assistant is behaving properly and correct it when it is not; and so that we have a reliable record of what was agreed, such as permission to check your insurance.

Recordings and written transcripts are stored securely. Only authorized staff can open them, and every time one is opened it is written to a log. If you would rather your call were not recorded, tell us and we will call you back on a line that is not.

What the AI collects, and how it is looked after

In a normal call or chat, the assistant collects your name, the phone number to call you back on, and the reason you are getting in touch. It does not go hunting for more than that.

If you ask us to check your insurance, it will also ask for your date of birth, your insurance company, and your member number, because a check cannot be run without them. Every one of these is optional. If you would rather not give one, say so — the assistant will drop it and our staff will verify your benefits by hand instead.

We also keep the call recording, the written transcript, and the chat conversation, along with the time and date.

All of this is treated as health information and handled under our HIPAA privacy practices — the same protections, the same access limits, and the same audit logging as any other record in this practice. Our Notice of Privacy Practices sets out what we may do with it and what rights you have over it.

We do not sell any of it, and we do not use it for advertising.

The AI can get things wrong

This technology is useful, but it is not perfect. It can mishear a word, misunderstand a question, or give an answer that is out of date or simply wrong. It can also sound confident while doing so.

For that reason, nothing the assistant says is a decision by this practice. It cannot commit us to an appointment, a price, a coverage amount, or a clinical opinion.

Messages taken by the assistant are read by a member of our staff before anything is acted on, and staff review calls and chats to catch mistakes and correct the assistant. If something it told you matters to you — a cost, a date, what to bring, whether we can help with your particular question — please ring the office and have a person confirm it.

If the assistant gets something wrong, or you are unhappy with the way it handled your call, please tell us on (631) 366-3369 and ask for the Office Manager. We want to know.

Why we publish this

New York has set out how it expects AI systems that hold conversations with people to behave (General Business Law article 47, section 1700 and following): tell the person clearly and conspicuously that they are dealing with an AI system rather than a human, and keep safeguards in place, including putting suicide-prevention resources in front of anyone who shows signs of self-harm. Federal and state consumer-protection regulators take a similar view of automated systems that could be mistaken for people.

We have written this practice to that standard, whether or not every part of it is aimed at a practice like ours. Our assistant identifies itself as automated at the start of every call, chat, and text message; it never claims to be a person or a clinician; it discloses that the call is recorded before recording anything you say; and it breaks off and gives the emergency instruction if someone describes a crisis.

This page is our own policy and disclosure. It has not been reviewed, certified, endorsed, or approved by any regulator, and nothing on it is legal advice or a statement of what the law requires of anyone else.

If you are experiencing a medical emergency, call 911 or go to the nearest emergency department. This practice does not provide emergency or crisis services.How We Use AI — Neuropsychological Associates