Indonesia's First AI-Animated TV Series Exposes Production Trade-offs
TRANS7 launched Legenda Bertuah in January 2026, Indonesia's first fully AI-animated television series. The 14 half-hour episodes adapt classic Nusantara folktales including Malin Kundang, Sangkuriang, and Timun Mas using generative AI for scripting, narration, and visual production, with human curators refining outputs.
The series reaches a core creative challenge: generative animation can scale cultural storytelling fast, but the technical limits remain visible to audiences. Viewers praised the show for making folktales accessible to younger viewers and noted cinematic visuals. Others flagged consistent problems - misaligned eye gaze, frame discontinuity, stiff expressions - that signal where current generative video tools still struggle.
How the production worked
TRANS7's workflow ran AI end-to-end: concept generation, script drafting, voice narration, and frame-by-frame visual synthesis. Prompt engineers then curated and sequenced the outputs, but could not fully correct motion coherence or facial performance realism at production scale.
The technical artifacts viewers noticed - jerky transitions, inconsistent gaze direction, abrupt frame changes - reflect known limits in temporal coherence and facial microexpression modeling. Generative video systems struggle to maintain consistent character behavior across sequences. No amount of prompt refinement fixes that gap at volume.
What creatives need to know
This deployment shows both the speed advantage and the quality cost of full automation. TRANS7 produced 14 episodes faster and cheaper than traditional animation would allow. The trade-off is craft: nuanced human performance, subtle expression shifts, and temporal smoothness remain difficult for generative tools.
For AI design work in animation and broadcast, Legenda Bertuah surfaces three practical problems. First, temporal jitter and facial artifacts become noticeable in longer narrative formats, even when individual frames look acceptable. Second, human-in-the-loop curation works for concept and direction but cannot scale to fix micro-level coherence issues. Third, the speed gains come with audience perception risks - viewers recognize AI artifacts, which affects credibility.
The broader questions
Critics and media commentators raised concerns beyond technical limits. Training data provenance remains undisclosed. The labor impact on animators and voice actors is real. Indonesia has no regulatory framework for AI-generated broadcast content, leaving questions about disclosure, copyright, and creator rights unanswered.
Public response split along practical lines. One viewer said: "It's really cool, quite refreshing and insightful." Another noted: "The expressions and tone sound very stiff, it's obvious that it's AI. That's one of AI's weaknesses: human performance can't be replaced by AI."
What happens next
Watch whether TRANS7 discloses model provenance and datasets. Monitor how Indonesian creative guilds and regulators respond. If TRANS7 pursues a second season, note whether they shift toward hybrid pipelines - using AI generation for backgrounds, layouts, or rough animation, then adding human refinement for character work and expression.
Improvements in temporal consistency and expression modeling will determine whether this becomes standard practice or remains a cost-driven exception. For now, Legenda Bertuah demonstrates that generative animation can handle folklore adaptation at scale, but quality-sensitive genres requiring nuanced human performance remain out of reach.
Your membership also unlocks:
