Open the app
ServicesAIMessagingPricingBlog Open the app

Privacy Policy

Lanterlink LLC · California · Last updated 19 November 2026

This policy covers two different things: this website, and the Lanterlink Care Portal that agencies log into. Our role is different in each, so we describe them separately.

The short version. On this website we collect what you type into a form, and we use it to reply to you or, if you sign up, to set up and bill your account. Inside the portal, the records belong to the agency that issued your login, and Lanterlink is that agency’s HIPAA business associate. The portal opens an agency’s records only after the agency has signed an agreement covering health information. Test accounts are for made-up information only. We do not sell personal information anywhere.

1. This website

If you request a walkthrough or ask about pricing, we collect the details you enter — typically your name, email address, agency name, and anything you write in a message field. We use that information to reply to you and, if it looks like a fit, to follow up once.

Signing up and paying. When an agency signs up on this website, we ask for the agency’s name, your name, role, work email and phone, its Regional Center and vendor number, roughly how many people it supports, any agency that referred it, and the services and counts it is paying for. We use these to check the sign-up, set up the agency’s workspace, and bill the subscription.

Payment is taken on a page run by Stripe, our payment processor, not on this website. The sign-up details are sent to Stripe with the order and kept in our Stripe account; this website does not keep its own copy. You enter your card and billing address on Stripe’s page. Lanterlink never sees or stores the full card number. Our Stripe account shows us only details such as the card type, its last four digits and its expiry date. We send Stripe no health information.

Stripe also uses payment information for its own purposes, including preventing fraud. On its payment page, that includes information about your device and browser. Stripe explains this in its privacy policy at stripe.com/privacy. Stripe’s code does not run on the other pages of this website.

Our web host records standard server logs, which include IP address, browser type, and the time of the request. These are used to keep the site available and secure.

We also count visits to this website using our web host’s analytics, which records the pages viewed, the site you came from, and your country and device type. It does not use cookies and does not identify individual visitors. If you use the search box on our blog, the same analytics also records the words you searched for and how many articles matched, so we can see which topics to write about next. It is not tied to you; please do not type names or other personal details into the search box.

We do not sell your information, and we do not add you to a mailing list because you contacted us.

Pages on this website load their typefaces from Google, which receives the visitor’s IP address when a page loads. We do not use that for advertising, and we do not let other companies track visitors across other websites. Stripe’s payment page is Stripe’s own: Stripe uses what it sees there, across the businesses it serves, to prevent fraud.

2. The Lanterlink Care Portal

The portal is used by staff at agencies that have a subscription with us. If you have a login, the agency that issued it — not Lanterlink — decides what is recorded about the people it supports and how long it is kept.

The portal opens an agency’s records only after the agency has signed a written agreement with us covering health information (section 4 explains it). Test accounts, which we set up for agencies trying the portal, are the one exception: they skip the agreement, they are for made-up information only, and the portal reminds their staff never to enter a real person’s details. Being a Regional Center vendor does not by itself decide the agency’s legal status. When the agency is a covered entity, or is itself handling health information for one, Lanterlink’s role is the one in that signed agreement, and the agreement controls how we may use the information.

3. What the portal holds

4. HIPAA

HIPAA is the federal law that protects health information. An agency bound by it generally may let a software company hold that information only under a written business associate agreement. Lanterlink is that business associate. Where the agency is itself a business associate — for example, of its Regional Center — Lanterlink is its subcontractor, and HIPAA treats a subcontractor as a business associate too. Whether an agency is also a covered entity in its own right is for the agency and its advisers to decide; our role is the same either way. HIPAA’s rules for business associates apply to Lanterlink directly, not only through the agreement.

Some of what an agency keeps is protected by California’s medical-privacy and developmental-services confidentiality laws rather than by HIPAA. The agreement extends its confidentiality, security, incident-reporting and return-or-destroy protections to that information as well.

Where we stand. In the business associate agreement, Lanterlink commits to meeting HIPAA’s rules for a business associate, including its security rule. HIPAA does not require a company to be certified, and the U.S. Department of Health and Human Services, which enforces HIPAA, says it does not endorse or recognize private “certifications.” So we show no “HIPAA certified” badge. Instead, this section says what we actually do, and an agency can ask us for the detail behind any of it.

The agreement comes first. An agency’s administrator signs the business associate agreement inside the portal before the portal opens the agency’s records (test accounts aside, as section 2 explains). Under it, we may use health information only to provide the service, for the limited purposes the agreement allows, or as the law requires. We do not sell it, use it for marketing or advertising, or use it to train AI models. On health information, an agency’s signed agreement wins over this page, and a change to this page cannot reduce what the agreement promises.

How we protect it.

Our providers. The outside companies that handle health information for us — Supabase, Google Cloud and Anthropic — each work under their own business associate agreement with Lanterlink, and the agreement an agency signs requires one before any provider handles its health information. Our subprocessors page lists them, with the date of each agreement.

If something goes wrong. If we discover a security incident, a breach, or any use or disclosure of health information that the agreement does not allow, we tell the agency without unreasonable delay and no later than 24 hours after we discover it. That first report says what we know at that point and does not wait for our investigation to finish. For a breach, the full details the agency needs follow as we learn them, and no later than 30 days after we discovered it. The only reason we may hold a report back is a delay that law enforcement is legally allowed to require. The agency coordinates the notices the law requires — to the people affected, to regulators, and to its Regional Center or other organization where that applies — and we help. Routine attempts that fail, such as blocked scans and failed sign-ins, are not reported one by one; we summarize them on request. If an agency suspects a problem with one of its own logins or devices, it should tell us without delay so we can help contain it.

People’s rights, and leaving. When someone asks the agency to see, copy or correct their records, or for a list of who their information was shared with, we help the agency answer in time (section 7). When an agency’s service ends, it can take its records with it, and we then return or destroy the health information we hold, including copies held by our providers, as the agreement sets out.

Questions. Write to privacy@lanterlink.com. An agency’s formal notices under its agreement, in either direction, go to the contacts named in that agreement.

5. How it is used

To operate the portal, to meet the reporting obligations agencies have to their funders and regulators — visit verification, Regional Center billing, payroll tax — and to answer support requests. We do not sell personal information, and we do not use the health information in the portal to advertise anything. The companies that store or process information from this website or the portal are listed on our subprocessors page. The typeface service described in section 1 is the one exception.

6. The optional AI tools

The portal includes optional AI tools for staff. Assist, Check and Translate help with a note. Ask Lanterlink is a staff assistant that answers questions about the agency’s own schedules and records. Voice lets staff dictate a note or speak to the assistant. A manager can also ask for a draft summary of a person’s notes for their 90-day report.

The agency decides whether to use them. They are on when a new agency joins, and its administrator can turn them off at any time in Settings, with nothing else in the portal affected. Agencies that joined before 2 November 2026 have them off until they turn them on.

An agency can also mark any person it supports as choosing no AI. The AI tools then do not work on that person’s notes, Ask Lanterlink leaves that person’s records out of its answers, and staff write those notes themselves. If someone else’s note or question mentions that person, their name is hidden from the AI.

What is sent. Each request carries only what the task needs:

Before anything goes to the AI, names, phone numbers, email addresses, ID numbers and street addresses are replaced with placeholders. The real values are put back when the answer returns. This works on patterns and cannot catch everything, so staff are asked to leave out unusual identifying details. Voice is the exception: a recording reaches Google’s speech service as sound, before any name in it can be hidden.

Who receives it. The AI is Claude, from Anthropic. Speech is turned into text, and text into speech, by Google Cloud. Both work under signed agreements covering health information, and both are listed on our subprocessors page. Anthropic may not use what we send to train its models. Google’s optional program that would let it keep audio to improve its service is not turned on.

How long they keep it. Anthropic deletes what we send, and what it sends back, within 30 days. If its automated safety system flags a request, it may keep that request for up to 2 years. Google turns a recording into text and does not keep the recording.

Where. We ask Anthropic to process requests in the United States only, and voice uses Google’s United States service.

A person decides. Every AI result is a suggestion. A note changes only when the staff member chooses to use the suggestion, and nothing is saved until they save the note. Ask Lanterlink cannot save, approve, sign, file or send anything by itself: when it prepares a form or a change, nothing happens until a staff member checks it and presses that screen’s own button. Dictated words land in the note box like typed words, for the staff member to read and save. AI can be wrong or leave things out, so staff should check every suggestion before relying on it. The AI tools are not used to write to or talk with the people an agency supports or their families, and they do not make care decisions.

What Lanterlink keeps. We do not store the questions, the answers, the text sent for a task, or the recordings, apart from what a staff member chooses to save in a note. A conversation with Ask Lanterlink lasts only while its window is open. We keep a count of AI requests (which tool, which month, and who made it) for 24 months, and then delete it; the agency uses it to see its use against its plan. Administrators see totals, and one person’s counts only in the limited cases shown in the agency’s AI settings. The counts are not a measure of anyone’s work. Each note also records whether AI helped write it.

Telling people. The agency is responsible for telling the people it supports about the AI tools, and we provide a printable notice it can hand out.

7. Your rights

People have the right to see, correct, or receive a copy of information held about them. Which path applies depends on whose information it is.

If you are portal staff, use the My Data button to download your own work records, or contact your agency administrator. A request about a person the agency supports goes to that agency. The agency holds those records and decides how the request is handled, and Lanterlink helps the agency meet the deadline that applies to the agency.

For anything relating to this website rather than the portal, email privacy@lanterlink.com.

8. How long we keep it

Records inside the portal are kept for the longest period that applies to that record. Some rows below are a legal minimum. Some are a longer hold we chose. The agency that issued your login can set a longer policy on top of this. If a period changes, we update this page.

RecordKept forWhy
Care and medical records7 years after the person leaves, and longer when the person was a minor: at least a year after they turn 18, and never under 7 yearsOur hold. A shorter period only if it is clearly lawful for that agency
Billing claims7 yearsOur own hold in the deletion tool, longer than the Regional Center minimum in the next row
Financial and service records tied to funding5 years from final payment for that state fiscal year, and longer while any audit or appeal is openCalifornia Regional Center record-keeping rules
Security and access logsKept. They are not on a deletion clockNot a promise that every view of a record is logged
Pay records and paystubs4 yearsOur hold period
The in-app activity log (who changed what)400 daysAn operational convenience record, not the compliance audit trail above
Counts of AI requests (which tool, which month, who made it)24 monthsOur hold. The questions, answers and recordings themselves are not kept
Inquiries sent through this websiteOnly as long as needed to answer them—

Deletion. An agency can ask us to delete its data earlier, and we will, except where a hold above still requires the record to exist. Before anything is deleted, our tool checks the record against the holds it knows about — billing history, service authorizations, and pay records — and will not proceed until an administrator reviews each warning and states a reason. That decision is written down, so nothing quietly disappears. Security and access logs are not deleted. We do not keep other customer data indefinitely by default.

9. Security

Access to the portal is restricted by role and by agency, changes to records are logged on the server, and data is encrypted in transit and at rest by the providers that store it. Those providers are under signed agreements covering health information. No system is perfectly secure, and we do not claim otherwise.

10. Changes

If this policy changes materially, we will update the date at the top and, for customers, say so in the portal.

11. Contact

Lanterlink LLC, California
Privacy: privacy@lanterlink.com
General: info@lanterlink.com

Lanterlink LLC · Built in California, for California providers.