Updated: October 25, 2026
AI has already shown up in your video quote
If you have asked a Miami production company for a quote in the last year, there is a good chance some part of the work involves AI tools, whether the proposal says so or not. Transcription, rough subtitles, noise cleanup, color matching, background removal and first-draft translations are now routinely done with machine learning. Some vendors go further and offer synthetic voiceovers, AI-generated B-roll, digital avatars of a spokesperson or automatic dubbing into Spanish and Portuguese.
Some of this saves real money. Some of it creates legal and brand risks that land on the client, not the vendor. This guide is for marketing managers and business owners buying video in Miami who want to understand what AI is doing in their project, how it should change the price, and which questions to ask before approving a quote or a final cut. It is practical guidance, not legal advice. For anything high stakes, involve your own counsel.
Where AI usually appears in a production
It helps to separate AI that assists human work from AI that produces the material your audience sees and hears.
Behind the scenes (low risk)
- Automatic transcripts of interviews so editors can find the best lines faster.
- Draft captions and subtitle timing, checked by a person before delivery.
- Audio cleanup: removing air conditioning hum, wind or crowd noise.
- Rotoscoping, object removal and upscaling in post-production.
- Rough first translations that a native speaker then corrects.
On screen (higher risk)
- Synthetic voiceover in place of a recorded voice actor.
- Voice cloning, where a real person's voice is recreated to say new lines.
- AI-generated footage, images or backgrounds that replace a shoot or stock clips.
- Digital avatars or face replacement of a real person.
- Automated dubbing and lip-sync into other languages.
The first group is normal production practice and should simply make the project faster or cheaper. The second group is where you need clear answers in writing.
The rights questions that matter most
Can you own what the machine made?
In the United States, copyright protects works of human authorship. The U.S. Copyright Office has said that material generated entirely by AI, without enough human creative control, is not protected on its own, while a work that combines human creative choices with AI-assisted parts can be. In practice, a video edited, scripted and directed by people with some AI-generated shots inside it is still a protectable work, but the purely generated pieces may not be yours to stop others from copying. If a logo animation, a mascot or a signature visual is central to your brand, ask whether it was generated, and consider having it made by a human designer.
Whose voice and face is it?
Cloning the voice or likeness of a real person needs that person's clear, written permission, and it should state exactly what the clone may be used for and for how long. That applies to your CEO, your employees and any hired talent. Union performers are covered by SAG-AFTRA agreements that address digital replicas, and non-union talent should still have a release that names synthetic use specifically. A standard on-camera release from five years ago almost certainly does not cover it. Our article on music, stock and talent buyouts explains how usage terms work for talent in general.
What did the tool learn from?
Some generative tools have been the subject of lawsuits over the material they were trained on. You cannot audit that yourself, but you can ask the vendor which tools they use, whether those tools offer commercial use terms, and whether the vendor will stand behind the output if a claim arises. Ask the same question about stock-style footage generated on request.
Your materials going into AI tools
A less obvious risk is the reverse direction. Your unreleased product footage, internal interviews, scripts and brand assets may be uploaded into third-party AI services during editing. Some services use uploaded content to improve their models unless the account is set otherwise. If your project is confidential, ask the vendor to confirm that client material only goes into tools whose terms do not allow training on it, or into tools that run locally. This connects directly to the confidentiality terms in your production contract.
Disclosure: when the audience must be told
Major platforms now ask creators to label realistic synthetic or altered content. YouTube asks uploaders to disclose realistic altered or generated material, and Meta applies AI labels to some content on Facebook and Instagram. Rules change, so whoever publishes your video should check the current policy before posting. Political advertising has its own disclosure rules for AI content, including in Florida, and regulated industries such as healthcare, finance and legal services have their own advertising standards. A synthetic doctor or an AI voice giving financial guidance is a bad idea regardless of what is technically allowed.
Brand trust is the other side of disclosure. A Miami audience that discovers the friendly bilingual narrator on your website is a generated voice may feel misled, even if nothing illegal happened. Decide in advance how open you want to be.
What AI should do to the price
AI can lower costs in some places and raise them in others. Ask the vendor to show where it is used so you can judge whether the price reflects it.
| Line item | Possible effect | What to check |
|---|---|---|
| Transcription and captions | Usually cheaper or faster | A person still proofreads the final file |
| Voiceover | Synthetic voice can cost less than a voice actor | Commercial license, approvals, how it sounds in Spanish |
| Translation and dubbing | Faster first drafts | Native speaker review is included |
| Generated B-roll | May replace a pickup shoot or stock | Rights, realism, brand fit |
| Avatar or voice clone | Setup cost, then cheap updates | Written consent, term, who controls the model |
| Review and legal | Can add time | Who checks output for errors and claims |
A quote that uses synthetic voice and generated footage but costs the same as a full live production is worth questioning. So is one that is much cheaper because human review has been cut. Our guide to comparing quotes and the article on hidden costs explain how to line up different approaches fairly.
Questions to ask before you approve
- Which parts of this project will use AI tools, and which tools?
- Will any voice, face or performance be synthetic or cloned? Who consented?
- Will any footage or imagery be generated instead of filmed or licensed?
- Does the tool's license allow commercial use, and will you stand behind the output?
- Will our confidential material be uploaded to any service that can train on it?
- Who reviews translations and captions, and are they native speakers?
- Will you tell us before introducing AI into a deliverable we already approved?
How Miami Video Company uses AI
We use AI tools where they help our editors and producers work faster: transcripts, caption drafts, audio cleanup and some visual effects. Everything a viewer sees and hears is checked by a person, and every translation is reviewed by a native speaker. We do not use synthetic voices, cloned voices, generated people or generated footage in a client project unless the client asks for it, approves it in writing and has the consents in place. When we do use them, we say so in the proposal and price them as separate line items.
If you are weighing whether AI makes sense for your next project, or you have a quote from another vendor that you want to understand, send it our way with your questions. We will tell you what the tools can do well, where human work still matters, and what each option would cost.