AI Video for Therapists

AI video can help a private practice share office logistics, general psychoeducation, and service descriptions. It is not a clinical tool. It cannot safely carry diagnosis, treatment advice, outcome promises, patient stories, protected health information, personalized follow-up, or crisis content. Those uses need strict clinical, ethics, privacy, and distribution gates that a marketing video cannot satisfy.

This article covers general marketing and education only. It is not mental-health advice, clinical guidance, supervision, or a substitute for your licensing board’s rules. Always consult your board, a qualified attorney, and your professional liability resource before publishing any practice video.

Where AI video fits

A short, warm video can answer quiet questions before a client books. It can explain what a first session looks like. It can describe your general approach or share broad well-being concepts. These are public-facing logistics and education, not treatment.

The moment a video implies a health benefit, shares a client story, or addresses a specific person's situation, the risk changes. The FTC health-ad guide says health claims must be truthful, not misleading, and supported before release. Testimonials and practitioner observations are not substitutes for sound evidence. A disclosure alone cannot fix a misleading net impression. For covered entities and business associates, HHS marketing guidance controls uses and disclosures of PHI for marketing. Written authorization is often required, subject to the rule's scope and exceptions.

A safer path is to build content in tiers. Office logistics and general psychoeducation sit at the lower-risk end. Service descriptions that explain your approach without promising outcomes can work when reviewed carefully. Health claims, outcome statements, testimonials, patient stories, patient-specific communication, and crisis or high-distress topics belong behind strict gates that a marketing video alone cannot open.

A content record for every video

Before you publish, document what the video contains and who approved it. A simple record can include:

• Intended audience
• Exact words and visuals
• Any express or implied claim
• The evidence that supports each claim
• Name and credentials of the licensed reviewer
• Privacy and authorization check, including a PHI stop gate
• Jurisdiction and board rule review
• Crisis-routing decision, if the topic touches distress
• Approval date, expiry date, archive location, and takedown owner

This record does not make a video safe by itself. It forces a deliberate review that can catch problems early.

A fictional first-appointment logistics video

Imagine a therapist who wants to explain the first-appointment process. They write a script with placeholders for arrival, check-in, forms, and the next step. It uses no diagnosis, treatment advice, outcome language, or real client detail. The practice reviews the exact script and distribution path before making the video.

This is a logistics video. It stays on the safe side of the line because it describes process, not care. If the same script added “most clients feel relief after the first session,” it would become a health claim and need a different level of support and review.

The PHI stop gate

Protected health information has no place in a public marketing video. Before you write a script or select a visual, pause and check for:

• Client names, initials, or identifying details
• Appointment dates, times, or schedules
• Screenshots of messages, records, or portals
• Lists of conditions, medications, or treatment plans
• Audio or images from real sessions
• Any story that could identify a past or current client

If any of these appear, stop. Rewrite the script so it speaks only to broad groups and common ideas. Never tailor public content to one person's case. The HHS mental-health privacy resources stress the sensitive nature of this information. Use a strict stop gate before any text, image, sound, or list enters an external tool.

Crisis and high-distress topics

If a video touches anxiety, trauma, grief, or similar topics, it can stir distress. A marketing video is not a crisis intervention. For these topics, require a current qualified protocol and locale-specific resources before you publish. Do not hard-code a single universal hotline and assume it fits every viewer. The protocol should name who reviewed the content, what resources appear, and how often they are rechecked.

Frame every frame as general education, not as a solution. Use inclusive language. Remind viewers the video is not a substitute for one-on-one care. At the start or end, point to the resources your protocol requires.

What Kyndrify documents

Kyndrify is one tool that can generate video from a user-supplied script. It documents a verified face-and-voice Twin built from a headshot and a short voice clip. Each Render produces a download and a hosted link. Longer scripts can be split and stitched.

The platform documents recorded consent, shared-credit billing, and AI disclosure. Signed Content Credentials and invisible provenance are rolling out and are not guaranteed on every file. See the Kyndrify pricing page and the Kyndrify Responsible AI page for current details.

C2PA Content Credentials and AI disclosure show that a video was AI-assisted. They do not prove clinical accuracy, consent for treatment, privacy compliance, safety, or ethics board approval. Those responsibilities stay with the practice.

Kyndrify’s public pages do not document rendering speed, templates, commercial rights, therapy-specific safeguards, integrations, analytics, or clinical outcomes. Current pages do document a Free tier with 8,400 one-time signup credits, Plus own-face and own-voice video up to 1080p, ten speech and dubbing languages, and a no-training statement for customer photos, voice, scripts, and finished media. Any claim beyond what the official pages document should be verified directly with the vendor.

Cost inputs for a practice video

A practice can estimate direct costs with a few inputs. These are illustrations, not promises. Check current pricing on the vendor’s official page before you commit.

• Script review: time for a licensed clinician to review the script, check board rules, and sign off.
• Consent recording: one-time setup to create a face-and-voice Twin.
• Render credits: each video Render uses credits from a shared balance. Subscription credits reset monthly and do not roll over, and pay-as-you-go packs remain usable for 3 months from purchase.
• Hosting and distribution: any cost to host the video file or embed it on your site.
• Periodic re-review: scheduled checks to confirm the video still meets current rules and resource links still work.

For a simple estimate, multiply the planned Render count by the current vendor credit rate. Add licensed review, privacy review, captions, hosting, monitoring, and update work. This gives a rough input-based figure, not a promised price or saving. Recheck the vendor's current official terms before approval.

A decision framework for therapists

Before you create any AI video, walk through these steps:

1. Define the tier. Is this office logistics, general psychoeducation, a service description, a health claim, a testimonial, patient-specific communication, or crisis content? The last four tiers need gates that a marketing video cannot reliably provide.
2. Write the script. Keep it general. Avoid PHI, outcomes, and personalized advice.
3. Check claims. List every express and implied claim. Match each to competent and reliable evidence.
4. Apply the PHI stop gate. Remove any protected health information.
5. Add crisis routing if needed. If the topic touches distress, include a reviewed protocol and locale-specific resources.
6. Obtain licensed review. A qualified professional should approve the script and final video.
7. Verify board and jurisdiction rules. Rules vary. What is permitted in one region may not be permitted in another.
8. Document the content record. Audience, claims, evidence, reviewer, privacy check, crisis decision, approval, expiry, archive, and takedown owner.
9. Generate and disclose. Use a tool that adds clear AI disclosure. Remember that disclosure does not replace the steps above.
10. Publish and schedule re-review. Set a calendar reminder to check the video’s accuracy and resource links.

Compact comparison: AI video vs. traditional filming

Factor	Traditional filming	AI video with your own Twin
Setup	Studio, lights, multiple takes	One-time consent recording
Script changes	Reshoot required	New Render from updated script
Update path	A new recording and edit may be needed	A new Render from the approved script is needed
Face and voice	Yours, recorded live	Yours, from a verified Twin
AI disclosure	Depends on the media and rules	Depends on the media and rules; verify the output
Privacy control	Depends on production workflow	Depends on platform data practices
Cost structure	Upfront production cost	Shared-credit billing; subscription credits reset monthly and do not roll over, and pay-as-you-go packs last 3 months

This table compares workflows, not outcomes. It does not imply that one method is safer or more effective for therapy marketing. The same ethical and legal gates apply to both.

A plain-language practice-video check

Use this list for each cut. Stop when a line fails.

• Name the public goal. Name who may see the clip.
• Mark the content tier. Do not call a high-risk cut low risk.
• Keep diagnosis out. Keep a care plan out. Keep advice broad.
• Mark each health claim. Mark what the art may imply.
• Link each claim to proof. Ask the licensed lead to check it.
• Mark each client story. Stop if a real person may be known.
• Remove each name. Remove each date. Remove each portal view.
• Remove each voice clip. Remove each face. Remove each case clue.
• Check the email list. Check the post list. Check each audience rule.
• Ask the privacy lead to review. Log the rule and any exception.
• Ask the board or ethics lead to review. Keep the reply.
• Check high-distress themes. Use the current local protocol.
• Check each help link. Name who will test it and when.
• Read the script as a new viewer. Note the full message.
• Watch the art with no sound. Check what it seems to promise.
• Hear the sound with no art. Check each claim and cue.
• Review the title and thumb. Treat them as part of the ad.
• Review the host text and links. Treat the page as one whole.
• Add true captions. Add each needed notice. Keep them clear.
• Add the AI use note. It does not prove clinical safety.
• Name the licensed signer. Name the privacy signer. Log both dates.
• Set the last-use date. Name who can pull the clip.
• Keep the script, proof, consent, sign-off, and live file.
• Pull a stale cut at once. Do not patch just one claim.
• Stop when a viewer may think this is care. Rewrite the frame.
• Stop when a phrase may sound like a cure. Cut it now.
• Stop when the help path is old. Fix it before launch.
• Stop when a client may be known. Strip each small clue.
• Stop when the board rule is not clear. Ask the right lead.
• Stop when the proof does not fit the claim. Drop the claim.
• Stop when the live page has changed. Review the whole page.
• Stop when a reply asks for care. Use the firm response plan.
• Stop when the clip may cause harm. Do not post it.

Frequently asked questions

Can I use AI video to explain my therapeutic approach? You can describe your general approach at a high level, such as what a term like “cognitive behavioral” means in your practice. Stop short of promising any outcome. Have a licensed reviewer check the script and final video. Follow your board’s advertising rules.

What makes a video a health claim? A health claim suggests that a service, method, or product can affect a condition, symptom, or state of well-being. Phrases like “reduces anxiety” or “improves relationships” are claims. They need competent and reliable scientific evidence. If you are unsure whether your script makes a claim, ask a qualified reviewer before you publish.

Does AI disclosure satisfy ethics requirements? AI disclosure is one part of transparency. It tells viewers the video was AI-assisted. It does not replace consent, privacy safeguards, board compliance, or clinical review. A disclosed video can still be misleading if the content itself makes unsupported claims.

How do I handle viewer comments or messages after a video? A public video can prompt personal replies. Never offer clinical advice in comments or direct messages. Have a policy for how you respond. If someone shares personal health information in a public comment, follow your privacy and record-keeping protocols.

What if my board has not issued AI-specific guidance? Apply the general rules your board already enforces for advertising, testimonials, confidentiality, and informed consent. When in doubt, consult your board directly or seek advice from a professional liability resource. The absence of AI-specific rules does not mean the existing rules do not apply.

Related reading

• AI Video for Chiropractors
• AI Video for Dentists
• How AI Video Short Form Can Boost Your Social Media

This article offers general information only. It is not mental-health advice, clinical guidance, legal advice, or a substitute for your licensing board’s rules. Pricing and features reflect a snapshot as of late September 2026. Always recheck the vendor’s official pages for current details before making decisions.