Every production company now has an opinion about AI, and most of them are marketing. One camp claims AI changes nothing. The other claims it changes everything. Neither is a useful answer if you are commissioning a video this year and simply want to know what the tools do, what they cost you in risk, and what you should ask your production company.
We are Fine Line Production in Beirut, and our position is easy to state because it is already published on our services page: we use AI, and we would rather say so plainly. This article is the longer version of that sentence.
How common is AI in video work now?
Very, and it is accelerating. In Wyzowl's Video Marketing Statistics 2026, an industry survey running since 2015, 63% of video marketers say they have used AI video tools to help create or edit marketing videos, up from 51% a year earlier. Whatever any individual company says about AI, the industry around it is clearly adopting the tools.
Which makes the real question not whether your production company uses AI, but whether it will tell you the truth about how.
Where AI actually helps
In our day-to-day work, AI is a workflow tool. We lean on it for proofreading, brainstorming, visuals, audio and speeding up the workflow. In animation projects, AI can carry look development and in-betweens, the connecting frames between key poses, while direction, script and edit stay in-house.
What these uses have in common: they compress the distance between an idea and its first draft. What they do not do is decide anything. The ideas, the craft and the final call stay human, and if a client prefers a fully AI-free process, they only have to say so and we deliver exactly that.
When AI is the brief itself
There is also a different kind of project, where generative AI is not a helper but the medium. We produce fully AI-generated films end to end: concept, generation, edit and sound, finished and delivered like any other Fine Line production.
The clearest example is Never Hide, a spec film we made to test how far generative video can carry a premium brand aesthetic. Cast, wardrobe, sets and grade were all prompted, then cut and finished in-house. Spec means it was our own exercise, not a client commission, and we label it that way deliberately. You can watch it and the rest of the experiments on our AI samples page.
Being able to produce a film this way teaches you precisely where the medium's edges are, which brings us to the other half of the article.
Where humans stay, and why
Our documentary work is the reason we are not tempted to oversell AI. When we film demining teams in Mount Lebanon, the value of the footage is that it is real: a real procedure, filmed at the pace it is actually done, with testimony from people who were injured by mines. A generated image of a deminer would not just be weaker, it would be a lie in a film whose entire purpose is evidence.
That is the honest boundary. AI is useful where the material is invented anyway, in brand worlds, animation and look development. It has no place standing in for reality in work that documents real people and real events. Most of our clients are UN agencies and NGOs whose films are trusted precisely because what is on screen happened.
Disclosure is becoming law, not courtesy
The regulatory direction is clear. Under the EU's AI Act, providers of generative AI have to ensure that AI-generated content is identifiable, and certain content, notably deepfakes, must be clearly and visibly labelled. The EU's transparency obligations take effect in August 2026. Even for organizations outside Europe, this sets the expectation international partners will work to, and many of the international organizations we work with answer to European funders and publics.
Our approach has been to get ahead of it by labelling honestly from the start: AI-generated work is presented as AI-generated, spec work is presented as spec, and documentary footage is what it claims to be.
What to ask your production company about AI
- "Where would you use AI on our project?" A trustworthy answer names specific steps, like drafts, look development or cleanup, rather than everything or nothing.
- "Will anything the audience sees be generated?" If yes, you should know what, and how it will be disclosed.
- "Can we have an AI-free process if we want one?" The answer should simply be yes.
- "Who makes the final call?" The only right answer is a person.
If you want to see what our answers look like in practice, the experiments are public on the AI samples page and the rest of the work is on our services page. Email info@finelineprod.com, call +961 1 741 744, or use the contact page.