The brand does not simply need another generated video
- A film must solve a brand or product task, but the brief begins with a format or technology.
- The team wants cinematic ambition at a budget or schedule that requires a different production route.
- Many short-form versions are needed without losing one recognisable creative system.
- AI experiments look impressive in isolation but cannot survive character consistency, product accuracy, rights or client approval.
- The project uses faces, voices, music, IP or data that require explicit permissions and records.
The cost of choosing technology before the communication task
Fast generation can produce more footage and more uncertainty. The film may have no narrative reason to exist, brand or product details may drift, revision becomes unpredictable and rights cannot be explained at release. Cheap tests become expensive when they must be rebuilt for quality, compliance or continuity.
The critical decisions
- What must the audience understand, feel or do after watching?
- Which format—commercial, brand film, explainer, short drama, series or long-form development—best serves that task?
- Which route should be live action, AI-generated, CG, animation or hybrid?
- What quality threshold, continuity, product accuracy and version system are required?
- Which likeness, voice, music, IP, dataset, generation and disclosure obligations must be controlled?
What the client buys
A visual narrative that begins with the brand task
The premise, script, characters and visual world are judged first by what they make the audience understand and remember.
A production route appropriate to the project
Technology is selected scene by scene according to creative value, quality, schedule, cost, consistency and rights—not because every project must use AI.
Release-ready multi-version assets
Masters, aspect ratios, durations, subtitles, cut-downs and platform versions are planned before production rather than improvised at delivery.
An auditable rights and generation record
Source material, permissions, releases, music, voices, likeness, tools, generation steps and disclosure requirements are recorded within the agreed scope.
The Meantime AI Film Accountability Chain
Insight
Clarify the brand task, audience, viewing situation, product truth, production constraints, reference risks and rights requirements.
Decide
Choose the narrative, format, production route, quality threshold, version matrix, budget priorities and approval gates.
Design
Develop concept, script, storyboard, look, characters, art direction, prompt and asset systems, sound and post-production approach.
Activate
Run casting or data preparation, shooting, generation, compositing, editing, sound, grading, subtitles, quality control and client reviews.
Review
Deliver masters and records, archive approved assets, document tool and rights information and evaluate audience response where distribution data is available.
Typical deliverables
- Film strategy, concept, script and storyboard.
- Visual development, style frames or proof-of-concept sequence.
- Commercials, brand films, explainers, short-form series and short dramas.
- Live-action, AI-generated, animation, CG or hybrid production.
- Editing, visual effects, sound, music, grading and subtitles.
- Multi-platform version matrix and master delivery.
- Releases, source, generation and rights documentation within scope.
Low-risk first engagement | Film concept and visual proof
For a team that needs to validate the creative and production route before committing a full budget. The first engagement produces a clear brand task, concept, selected script or treatment, representative visual frames or a short proof, feasibility findings and a costed next-stage route.
Measurement and acceptance
Acceptance covers approved creative documents, stated technical specifications, product and brand accuracy, continuity, sound and image quality, version delivery and agreed rights records. Distribution performance is assessed only when a media plan, baseline and data access are part of the scope. Views, virality or sales are not guaranteed.
Work and evidence boundaries
Portfolios distinguish concept tests, company-produced work, partner roles and released client projects. AI-assisted work identifies the relevant production route where disclosure is required. No film frame, celebrity likeness, voice, music, IP or client asset is treated as publishable merely because it exists in a project file.
Who this is for—and who it is not for
A good fit
- Brands that need film to perform a clear communication or growth task.
- Projects where AI, live action or a hybrid route should be evaluated responsibly.
- Teams requiring multi-version assets and a controlled review process.
- Clients prepared to clear rights and approve a defined quality route.
Not a good fit
- Bulk generation without a brand, audience or release standard.
- Use of faces, voices, music, IP or confidential data without permission.
- A guaranteed claim that AI will always be cheaper or faster.
- A deadline incompatible with creative approval, quality control or rights clearance.
Frequently asked questions
Does every project use AI?
No. The production route follows the creative and commercial task. Live action, animation, CG and hybrid methods remain valid choices.
Is AI always cheaper and faster?
No. It can accelerate some stages and create cost elsewhere through iteration, continuity, compositing, quality control or rights review.
Must AI-generated video be labelled?
Applicable platform, contractual and legal requirements are assessed for the actual release. The agreed project record states responsibilities; it is not legal advice.
Can we use a celebrity, employee or customer’s face and voice?
Only with appropriate authority for the intended use. Synthetic likeness and voice require the same disciplined rights analysis.
How are music, IP and generated assets cleared?
Sources, licences, client-supplied materials, tool terms and intended uses are documented. Specialist legal review may be required.
What files and revision rounds are delivered?
Masters, versions, working files, source assets and revision rounds are defined in the statement of work. Nothing is assumed from the format name alone.