This week, Netflix published its first comprehensive guidelines on the use of generative artificial intelligence (GenAI) in content production, outlining both acceptable applications and the conditions under which formal approval is required.

The Los Gatos-based streaming giant acknowledges GenAI’s growing role in creative workflows, particularly as a tool for ideation, and considers it a “valuable creative aid when used transparently and responsibly”. In practice, the company allows its use for non-final purposes such as moodboards, reference images and temporary mock-ups, provided production partners inform their Netflix contact in advance.

Beyond these low-risk cases, producers must escalate to Netflix for written approval. The guidance is structured around four key areas: data use, creative output, talent and performance, and ethics and representation.

The platform has set out five guiding principles, including the prohibition of replicating copyrighted or unowned material, the requirement to work in secure environments where possible, and the commitment not to replace or generate talent performances without explicit consent. Moreover, generated material should remain temporary and not appear in final deliverables unless approved.

On the data side, Netflix insists that no proprietary or personal information – such as unreleased scripts, production images or cast details – be entered into GenAI tools unless specifically cleared. Likewise, producers are instructed not to fine-tune models using third-party works without the necessary rights.

From a creative standpoint, approval is required whenever GenAI is used to generate key characters, central visual elements or fictional settings. The guidelines also caution against prompts referencing copyrighted artworks, deceased figures or public personalities.

Particular care must be taken in relation to performers. Creating digital replicas of real actors or making significant alterations to their performances requires documented consent and compliance with union regulations. Even subtle changes to delivery, tone or intent may raise legal and reputational concerns.

The ethical dimension is equally central: Netflix warns against producing misleading or misrepresentative content that could blur the line between fact and fiction, such as fabricated news footage. The company also stresses that GenAI should not replace union-covered roles, including those of actors, writers and crew members, without agreements in place.

To help producers navigate these scenarios, Netflix has introduced a “Use Case Matrix”, classifying common applications of GenAI according to their level of risk and the need for escalation. Custom workflows developed by vendors or AI studios are also subject to the same standards at every step of the pipeline.

The full guidelines are now available on Netflix’s Partner Help Center and can be viewed here.