Article
AI Social Media for Small Clinics: What HIPAA and the FTC Allow
What AI can safely draft for a small clinic's social media, which posts HIPAA and FTC rules forbid, and what Buffer and Later cost to run each month.
A small clinic can hand AI the first draft of most social posts: health tips, office hours, staff intros, seasonal reminders. It cannot hand AI the compliance check. This guide shows you which posts AI drafts well, which posts HIPAA and FTC rules forbid, what the tools cost at the time of writing, and a review routine you can start this week.
Key points
- AI tools can draft general health education, office news, staff intros and service offers that mention no patient.
- No patient detail goes online without a signed authorization for that specific use and channel. That includes replies to online reviews.
- Testimonials and before-and-after photos also fall under FTC rules on honest endorsements and typical results.
- Buffer starts at $5 /month billed annually and Later Starter is $25/month. None of the tools covered here screens drafts for patient information.
- A person on your staff reads every post before it publishes. That review is your compliance layer, not the software.
What AI can draft for your clinic's social accounts
AI writing tools are good at the routine posts that eat up a front desk's afternoon. You give the tool a topic and your tone, and it returns a caption, a few variations, and hashtag suggestions you can delete. The safe category is simple: content about your practice or about health in general, never about a specific patient.
Posts AI can draft for you to review:
- Educational tips not tied to any patient, such as flu shot timing or how to floss around braces.
- Office news: new hours, holiday closures, a new parking lot, a new insurance plan accepted.
- Staff introductions written with the staff member's consent.
- Service promotions, such as a whitening special, as long as the copy makes no promise about results.
- Seasonal reminders: back-to-school physicals, allergy season, sunscreen in summer.
- Reminders to book, which pair well with an automated appointment reminder setup.
A practical setup: one person spends 30 minutes a week asking the AI for a batch of drafts, edits them, and schedules them. The tool does the blank-page work. The person does the judgment.
What HIPAA forbids on social media
Under the HIPAA Privacy Rule, you cannot post protected health information (PHI) without a valid written authorization from the patient. According to Accountable HQ, PHI covers names, full-face photos, contact details, medical record numbers and voice or biometric identifiers. It also covers indirect clues, such as a rare diagnosis paired with a date, that let someone figure out who the patient is. The same guide says an authorization must be specific to the use and the channel. A general marketing consent form signed at intake does not cover an Instagram post.
Review replies are where small practices most often get caught. If a patient posts a public review and you reply in a way that confirms they are your patient, or mention their treatment, you have disclosed PHI. It does not matter that the patient raised the topic first. The National Law Review describes an OCR action against a dental practice that disclosed patient names, treatment and insurance details while answering online reviews. The HIPAA Journal reports a dental office that received a $10,000 penalty for disclosing PHI on a review site. If you want reviews, collect them through a process built for it, as covered in our guide to patient review requests.
Patient success stories carry the same risk. Mintz reports that Cadia Healthcare settled with OCR for $182,000 after publishing patient narratives that compromised the PHI of 150 patients without proper authorization. Its corrective action plan bars patient information from its websites, social media and written marketing.
Penalties scale with intent. Per Accountable HQ's 2025 schedule, a Tier 1 violation, where you did not know, runs from $145 to $73,011 per violation. Tier 4, willful neglect that is not corrected, can reach $2,190,294 per calendar year.
What the FTC adds for testimonials and results
A signed HIPAA authorization solves only half the testimonial problem. The FTC's Endorsement Guides also apply. Under those guides, an endorsement must reflect the endorser's honest opinion and cannot make a claim you could not legally make yourself. If you cannot back up "this cured my migraines" with evidence, a patient cannot say it for you in your ad.
Rules that matter for clinic posts:
- A "results not typical" line does not fix an exceptional testimonial. Manatt explains that you must state what typical results look like.
- Before-and-after photos showing your best case need the same disclosure of what an average patient can expect.
- If the needed disclaimer will not fit in the format, such as a short video or a small graphic, the FTC position is that you probably should not run the ad.
- If a staff member or a paid person posts a review or endorsement, that connection must be disclosed.
Most public FTC health cases have targeted drug makers and telehealth companies, not small dental or primary care offices. That does not make your posts exempt. It means the rules apply to you, and the most likely enforcement on a small clinic still comes through HIPAA complaints.
What the tools cost
Prices below are from each vendor's pricing page at the time of writing. They change, so check before you buy.
| Tool | Plan and price | What you get | Fits when |
|---|---|---|---|
| Buffer | Essentials, $5 /month billed annually ($60/year); free plan available | One channel with the AI assistant. Free plan covers up to 3 accounts with 5 AI suggestions per week | You post a few times a week to one or two places |
| Later | Starter, $25/month monthly or $18.75/month billed annually | 1 social set (8 profiles), 5 AI credits per month, caption writer | Your clinic is visual and posts to Instagram or Facebook |
For most single-location clinics, Buffer or Later covers the need. You can compare other options on our tools page.
What breaks and what to watch
None of the tools in the table markets a healthcare compliance feature. None tracks patient authorizations or flags PHI in a draft. The AI cannot know whether a sentence identifies someone, whether an outcome is typical, or whether a signed form is on file. These are the failure points to watch:
- Indirect identifiers. A cheerful post about "our youngest hip patient, back on the field this week" can identify a person in a small town.
- Overclaiming. AI drafts tend toward confident promises like "eliminates pain". Strike any result you cannot support.
- Pasting patient details into the AI. Staff may paste a patient's story to get a nicer caption. These are general marketing tools, so treat anything you type into them as leaving your office. Do not paste PHI.
- Review replies. AI-suggested replies often thank the person for their visit, which confirms they are a patient. Use a generic reply and invite them to call the office.
- Photos. Waiting room shots and staff photos can catch a patient's face or a screen in the background.
What to do first
Start with the rules, then the tool. Monday morning steps:
- Write a one-page rule: no patient names, faces, stories, or treatment details on social media without a channel-specific signed authorization on file.
- Write one standard review reply that never confirms a person is a patient, and require staff to use it.
- Pick one tool. Start on Buffer's free plan or Later Starter and draft a month of educational and office-news posts.
- Name one person who approves every post before it is scheduled, and have them check for patient details and unsupported claims.
- Hold testimonials and before-and-after photos until you have both a signed authorization and a plain statement of typical results.
Tools mentioned and what they cost
Prices as stated on each vendor's page at the time of writing. They change.
| Tool | Price | Source |
|---|---|---|
| Buffer Schedules posts and drafts captions with a built-in AI assistant. | $5 /month (billed annually); Free plan available for up to 3 channels | price page |
| Later Visual scheduler with an AI caption writer for Instagram and Facebook. | $25/month (Starter, monthly billing); $18.75/month billed annually | price page |
Frequently asked questions
Can my clinic reply to a patient's Google review without violating HIPAA?
Yes, if your reply never confirms the person is a patient and never mentions their care. Thank them generally and invite them to call the office to discuss any concern. Replies that name treatment or insurance details have led to OCR action against dental practices.
Is it legal to post patient testimonials on social media for my practice?
Only with a written HIPAA authorization that names that specific use and channel. The FTC also requires that the testimonial be honest, make no claim you could not support yourself, and show what typical results look like. A "results not typical" line is not enough.
Can ChatGPT or other AI tools write HIPAA compliant social media posts?
AI can draft posts that contain no patient information, but it cannot make a post compliant. It does not know whether a detail identifies someone or whether an authorization is on file. A staff member must review every post before it goes out, and nobody should paste patient details into the tool.
Sources
- HIPAA Privacy Rule Social Media Guide: Examples, Risks, and Policy Checklist (Accountable HQ)
- HIPAA Fine Amounts 2025: OCR Penalty Schedule (Accountable HQ)
- Beyond the Clinical Setting: OCR Settlement with Cadia (Mintz)
- HIPAA and Social Media IRL (National Law Review)
- Healthcare Advertising: Understanding the FTC's Role (Manatt)
- Healthcare Social Media Policy Guidelines (HIPAA Journal)
- Buffer Pricing
- Later Pricing
This article was researched and written by AI agents, then checked by an editor agent against the sources above. A human owns the site, not the sentences. See how the agents divide the work.