[ Standard / Evaluation ]

The 51% AI Participation Standard

Eligible AI-native films show AI as the leading creative force across weighted dimensions of the production—not a lone VFX statistic or incidental assist.

What 51%+ AI participation means

The threshold is an expert review of holistic creative dominance, not a spreadsheet cell or single render pass percentage.

  • Jurors synthesize documentary evidence—including creative journals, timelines, prompts, selects, mix stems, VFX turnovers, and graded masters—to judge where AI materially led each dimension.
  • The number refers to dominance across storytelling dimensions collectively. A film does not qualify by maxing one category while ignoring the rest—it must convincingly surpass the bar in aggregate.
  • Human authorship remains vital: directors and producers steer ethics, taste, and final approval. The standard simply clarifies where AI—not traditional departments alone—drives inventive outcomes.

Six creative dimensions evaluated

Each dimension earns a juried weight toward the whole. Together they sum to 100%.

Total weight100%
Writing (screenplay)20%
  • ·Story architecture shaped with AI-assisted ideation.
  • ·Character arcs refined through iterative AI drafts.
  • ·Dialogue experimentation beyond light proofreading.
Direction20%
  • ·Shot sequencing planned with AI visual references.
  • ·Performance tone explored via generative rehearsals.
  • ·Narrative blocking informed by diffusion or sim tools.
Generation (image / video)20%
  • ·Principal imagery created or heavily rebuilt with generative pipelines.
  • ·Look-dev cycles driven by diffusion / video models.
  • ·Synthetic environments that define on-screen reality.
Editing15%
  • ·Structural iterations generated or ranked with AI assistance.
  • ·Temp scores or rhythm studies created synthetically.
  • ·Alternate cuts explored at scale before picture lock.
Sound15%
  • ·Stem-level design co-created with generative audio tools.
  • ·Adaptive score palettes shaped by AI timbre explorers.
  • ·Soundscapes impossible to source solely from library foley.
Post (VFX · finishing)10%
  • ·Compositor-grade passes leaning on neural rendering.
  • ·Color decisions iterated through AI-assisted looks.
  • ·Mastering tweaks directly tied to generative rework.

How to self-assess your film

  • 01Would removing AI workflows materially change the story beats or emotional arc?
  • 02Did diffusion, video, or neural tools define the iconic visuals audiences remember?
  • 03Were editing timelines or pacing explored through AI-generated alternates?
  • 04Does the soundtrack hinge on synthetic instruments, generative ambience, or AI-assisted cues?
  • 05Did VFX finishing or grading rely on generative rework—not just cleanup?
  • 06Can you assemble dated evidence (journals, exports, renders) proving AI leadership per dimension?

What HKAIIFF will verify

  • Signed creative diary tying major decisions to AI explorations versus traditional plates.
  • Tool logs / render stamps / cloud job IDs showing model usage across departments.
  • Side-by-side evidence of AI output ratios blended into the locked master—for example stem splits or VFX turnovers.