Recognising the increasing use of generative AI across creative workflows, Netflix has produced guidelines aimed at informing filmmakers, production partners and vendors when and how to use artificial intelligence.
Citing industry-wide concerns around generative AI and the evolving legal landscape, the company said it expects all production partners to declare any intended use, as new tools continue to appear with different capabilities and risks.
Netflix has outlined five guiding principles to be considered before generative AI is leveraged in creative workflows. Ahead of any deployment, production partners should answer the following:
- The outputs do not replicate or substantially recreate identifiable characteristics of unowned or copyrighted material, or infringe any copyright-protected works
- The generative tools used do not store, reuse, or train on production data inputs or outputs.
- Where possible, generative tools are used in an enterprise-secured environment to safeguard inputs.
- Generated material is temporary and not part of the final deliverables.
- GenAI is not used to replace or generate new talent performances or union-covered work without consent.
If any of the questions are answered “no” or “unsure”, further guidance should be sought before proceeding, said Netflix. The same principles apply to partner vendors using custom gen AI workflows – pipelines built from multiple tools or models.
Written approval
Some cases will always require written approval. Examples include:
Data use. Creators should consider whether data used requires any special handling, clearance or consent. Netflix-owned materials (unreleased assets and materials) or personal data should not be input without explicit approval, said the company. Models should not be trained on unowned or third party talent assets without legal clearance.
Gen AI should not be used to generate main characters, key visual elements or settings without written approval. Use of copyrighted or estate-controlled materials should be avoided.
Synthetic or digital replicas should not be created without consent and compliance with appropriate guild requirements. Care should be taken when making significant alterations to performances that affect “emotional tone, delivery or intent,” said Netflix.
Misleading or misrepresentative content should be avoided to ensure reputational harm does not occur. Fabricated footage, dialogues or scenes should not be presented as authentic. Work done by union-represented individuals should not be replaced or materially impacted by gen AI.
If unsure, check
The guidance further outlines how to ensure content and data protection, advising partners to speak with their primary Netflix contacts about available tools and solutions. licensing tiers and usage.
All gen AI-created content that appears on screen—even in the background—can potentially raise legal, trust or copyright issues, which is why it should all be flagged at an early stage in production, said the company.
Extensive guidelines have been put in place for talent enhancement use cases, for which there is a long tradition on post production and VFX. Given the emerging legal, ethical and reputational challenges, it is essential that creators exercise caution at every stage of the process and that consent is placed front and centre of any consideration, added Netflix.
Vendors and production partners using custom workflows must ensure standards are met at each stage of the pipeline. Again, if there is any uncertainty, guidance must be sought.
Netflix has produced a proposed use case matrix as a tool to provide speedy guidance for partners. The guidelines aim to address some of the concerns around the technology and ensure legal, reputational and ethical standards are met as the use of generative AI expands across the media and entertainment environment.
