During project initiation, first confirm who is responsible for visual continuity in the AIGC commercial.

When launching an AIGC commercial project, brands most often overlook assigning responsibility for visual continuity. In traditional live-action productions, continuity is jointly ensured by the director, script supervisor, and editor, whereas in AIGC commercials, it is typically fragmented across generative models, post-production compositing, and manual retouching. Responsibilities for each continuity issue must be clearly defined during initiation; otherwise, blame-shifting will occur later.

AIGC commercial footage from case materials, observing camera angles, subjects, and lighting relationships.
Frame capture from case materials, sourced from the study 'Judged by AI, 90 minutes to prove innocence, the VFX of Mercy.' This image is for observing camera work and production methods only and does not represent an ONCE client project. Source page. Case Materials Page

Specifically, require the production team to submit a continuity management plan at the kickoff meeting, covering character appearance consistency, scene lighting consistency, camera movement coherence, and prop detail stability. The plan must specify generation parameters, seed values, or reference images for each shot, along with remediation protocols for inconsistencies. If the team cannot provide such a plan, project risk increases significantly.

The risk is that while many AIGC tools produce high-quality individual frames, sequential playback may reveal facial distortions, flickering clothing textures, and drifting background elements. Fixing these issues can cost more than regenerating the content. An exception applies if the ad is intended solely for social media stills or short looping videos with rapid cuts; in such cases, continuity requirements may be relaxed, but acceptance criteria must still be specified in the contract.

As a delivery consequence, if responsibilities are not clearly defined during project initiation, the brand may find during final acceptance that all issues are attributed to AI tool uncertainty, leaving the production team under no obligation to make corrections. Therefore, continuity acceptance clauses must be included in the contract or statement of work at project kickoff, and a technical lead within the production team must be designated.

The quality of AIGC-generated ads depends on the materials provided by the brand.

When supplying materials, brands should not provide only product images and a tagline. AIGC generation models require sufficient reference information to maintain visual consistency. Prepared materials should cover the following categories, each with specific quantity and quality requirements.

  • Brand visual guidelines, including logo, standard colors, typography, and supporting graphics; at minimum, vector files and color codes must be provided.
  • Multi-angle product photos are recommended, including front, side, back, and close-up detail shots, with at least three images per angle under different lighting conditions.
  • Visual references for the target audience, including images or video clips depicting age, gender, fashion style, and lifestyle settings.
  • References for film or commercial styles preferred by the brand, with specific titles or links and noted timestamps of favored moments.
  • A clear list of delivery platforms, such as Douyin, WeChat Channels, Bilibili, and TikTok; resolution, frame rate, and duration limits for each platform should follow the latest official specifications prior to publication.

The general rule is that more specific materials yield greater visual stability. If the brand provides only text descriptions, the production team must source its own reference images, increasing uncertainty. A key risk is that brand-supplied references containing third-party logos or likenesses may raise copyright issues, so the production team should review them before use.

An exception applies when a brand is in the creative exploration phase and uncertain about visual direction; in this case, providing only core selling points and emotional keywords allows the production team to generate multiple style proposals. However, visual continuity acceptance standards should be lowered at this stage, as outputs from style exploration are not intended for final delivery.

How to Reserve Room for Continuity Fixes During the Shooting Phase of AIGC Commercials

Even if most visuals in an AIGC commercial are generated by AI models, the shooting phase must still prepare assets for post-production compositing. Brands should require the production team to create a complete shot list and storyboard before filming or generation, specifying the production method for each shot, including pure AI generation, live-action with AI compositing, or live-action with CGI replacement.

Specifically, for live-action scenes, require the camera crew to use the same camera body, lens set, white balance, and color profile, and to capture gray and color charts at the start of each scene. For purely AI-generated scenes, require the production team to save generation parameters for every shot, including model version, prompt, seed value, sampling steps, and CFG scale, as these are critical for later reproduction or correction.

The standard requires retaining at least two generated versions per shot: one final version and one backup. The backup enables quick replacement if continuity issues arise in the final version. A key risk is that production teams may save only the final output without retaining generation parameters, preventing post-production adjustments.

An exception exists if the commercial uses real-time generation technology, such as LED backgrounds in virtual production, in which case frame sequences of generated imagery must be recorded synchronously during shooting for unified post-production review. Consequently, failing to save parameters and backup versions during shooting may force regeneration during post-production, increasing time and costs while potentially creating stylistic inconsistencies with approved shots.

Controlling Continuity in AIGC Commercials via Shot Matching Charts in Post-Production

Post-production is when continuity issues in AIGC commercials typically surface. During post-production review, brands should require the production team to maintain a shot matching chart documenting key attributes for each shot, including character facial features, costume colors, scene lighting direction, focal length, and motion paths. This chart serves as the basis for post-production corrections and final acceptance.

Specifically, after completing the rough cut, the production team should conduct a frame-by-frame continuity check between adjacent shots, focusing on the following items.

  1. Consistency of character facial features, including eye size, nose bridge height, skin tone, and hairstyle.
  2. Check continuity of costume details, including wrinkles, patterns, and button placement.
  3. Verify that prop placement within the scene remains stable, such as cups on tables or artwork on walls.
  4. Ensure lighting direction is consistent, including key light angle, shadow direction, and specular highlight position.
  5. Confirm camera movement continuity, including the speed and direction of zooms, pans, tilts, and dollies.

If any inconsistency occurs, the production team must flag and fix it using methods like inpainting, image repair, frame interpolation, or regeneration. However, fixes may introduce new discontinuities, such as noise mismatches between repaired areas and surrounding pixels. Therefore, a full-sequence preview is mandatory after repairs rather than reviewing single frames.

An exception applies to fast-cut commercials where shots last under 0.5 seconds; since viewers naturally overlook minor inconsistencies, acceptance criteria may be relaxed. However, the brand must explicitly communicate this stylistic preference to avoid unnecessary over-correction and increased post-production costs.

Without a shot-matching log delivered during post-production, the brand can only rely on subjective judgment during final acceptance and cannot quantify issues. Brands should contractually require the production team to deliver a shot-matching log along with repair records for each shot.

Copyright verification is a critical step distinguishing AIGC commercials from traditional advertising.

Copyright issues in AIGC commercials are more complex than in live-action productions because generative models may use datasets containing third-party works. During copyright acceptance, brands must secure usage rights not only for the final video but also verify legal authorization for all assets, models, and tools used in the generation process.

Specifically, request the following list of copyright documentation from the production team.

  • The name, version, and provider of the AIGC tools used, along with the provider's copyright policy for generated content.
  • Whether the training data includes copyrighted works and, if so, whether the provider has obtained authorization.
  • Sources and licensing proof for reference images, prompts, and style transfer assets used during generation.
  • Original files for live-action footage, talent likeness releases, and location permits.
  • Licensing documents for music, sound effects, fonts, and icons, particularly regarding commercial usage rights.

The criterion is that all licenses must cover the brand's intended usage, including platforms, territories, duration, and modification rights. The risk is that some AIGC tool terms restrict generated content to personal learning and prohibit commercial use, exposing brands to infringement liability if used directly. Therefore, brands should require production teams to provide a summary of tool terms at project initiation.

An exception applies if the brand plans overseas distribution of AIGC ads, requiring attention to target region copyright laws, such as the EU AI Act and U.S. Copyright Office policies. As these regulations may change, brands should rely on the latest official information.

The consequence is that incomplete copyright documentation may lead to third-party claims post-launch, with no recourse against the production team. Therefore, copyright acceptance should be a prerequisite for final payment, with the balance paid only after all licensing documents are received and approved.

The final deliverable acceptance checklist covers visual continuity, copyright clearance, and delivery formats.

Upon receiving the final AIGC ad, brands should use a structured acceptance checklist to verify and document each item. The checklist comprises three dimensions, each containing specific inspection items and pass criteria.

For visual continuity, check character consistency, scene consistency, lighting consistency, and motion coherence. The pass criteria require no noticeable jumps in these attributes during adjacent shot transitions, ensuring viewers detect no anomalies during playback.

For copyright, verify that all visual elements, audio, fonts, and generative tools in the final video are authorized for commercial use. The pass criteria require valid authorization documents for each item, with coverage extending to the actual distribution scenarios.

For delivery formats, check resolution, frame rate, codec, color space, subtitle files, master versions, and source file packages. The pass criteria require compliance with platform publishing standards and inclusion of all assets and project files to facilitate future edits.

Specifically, brands should use professional players for frame-by-frame inspection during acceptance rather than relying on compressed previews. For continuity issues, compare screenshots of adjacent frames using basic image editing software to zoom in on details. For copyright, request scanned or digital copies of authorization documents from the production team and verify the validity period and scope.

A key risk is brands focusing solely on the final video while neglecting delivery formats, resulting in unavailable source files for future edits. An exception applies if the brand only purchases usage rights to the final video and does not need source files; this should be specified in the contract, though pricing typically differs.

Identify scenarios unsuitable for AIGC commercials in advance.

Not all advertising projects are suitable for AIGC technology. Brands should evaluate the following factors during project initiation; if any apply, traditional live-action or hybrid production is recommended.

  1. Products requiring extremely precise color accuracy, such as cosmetic shades or automotive paint finishes, may suffer from color deviations in AIGC-generated imagery.
  2. Commercials needing to demonstrate authentic physical properties like texture, weight, or temperature are unsuitable, as AIGC cannot simulate real tactile sensations.
  3. Commercials featuring extensive dialogue or performance that rely on genuine actor expressions and emotions may be unsuitable, as AIGC-generated facial expressions can appear unnatural.
  4. Brands have extremely high requirements for visual continuity; for example, product demonstration videos require millimeter-level precision, and the randomness of AIGC may cause detail drift.
  5. When project timelines are extremely tight and budgets limited, AIGC iteration and correction can be more time-consuming than traditional live-action shooting.

The criterion is that if a brand's core selling points rely on realism or precision, AIGC commercials may not be suitable. The risk is that forcing AIGC use can lead to repeated rework, ultimately exceeding the cost of traditional production. An exception applies if the video is intended solely for concept presentation or internal proposals rather than final distribution, in which case AIGC can quickly generate visual references.

Consequently, if brands fail to identify unsuitable scenarios during project initiation, they may face missed deadlines or substandard quality later. Therefore, it is recommended that brands conduct a technical feasibility assessment with the production team before launch to discuss the applicable boundaries of AIGC.

The next step is to start with a small-scale test shot.

For brands evaluating AIGC commercials, the safest approach is to first produce a 10- to 15-second test shot covering a complete product showcase or scene transition. This test should address key continuity challenges, such as character turns, product close-ups, and scene transitions. Through this test, brands can directly assess AIGC visual quality, continuity performance, and copyright risks while evaluating the production team's communication and responsiveness.

Test shots cost far less than full projects yet provide sufficient data for decision-making. Based on actual output, brands can decide whether to proceed with the full project or adjust the production plan. If test results are unsatisfactory, brands can promptly pivot to traditional live-action or hybrid production to avoid greater investment.

Finally, regardless of the production method chosen, brands must actively verify final video quality and avoid lowering standards simply because the technology is novel. While the core value of AIGC commercials lies in efficiency and novelty, campaign success still depends on clear messaging, engaging visuals, and copyright safety.

If you are preparing an AIGC commercial project, gather your brief, visual references, product or corporate materials, delivery platforms, and copyright scope before reviewing ourAIGC Video Services page, translating abstract preferences into actionable production boundaries.