Netflix is addressing criticism concerning its use of artificial intelligence (AI) in the upcoming true crime documentary, What Jennifer Did, directed by Jenny Popplewell. The documentary faced scrutiny for seemingly integrating AI-generated images instead of actual archival photographs, raising concerns about the portrayal of reality in narratives where factual representation is paramount.
In response to this backlash, Netflix has issued new guidelines concerning generative AI usage in content production. The platform published a detailed post on its Partner Help Center outlining the principles that it expects production teams to adhere to when utilizing this technology. According to Netflix, generative AI tools can serve as “valuable creative aids,” facilitating the rapid creation of unique media formats ranging from video and sound to text and images.
The guidelines come at a time when developments in the generative AI realm are progressing rapidly, prompting the company to establish a clear framework for its partners. “To support global productions and align with best practices, we ask all production partners to communicate their intended use of generative AI with their Netflix contact, especially as new tools emerge with varying capabilities and risks,” the announcement states. It clarifies that while most low-risk applications adhering to these principles may not necessitate legal review, instances that involve final outputs, likeness of talent, personal data, or third-party intellectual property will require prior written approval.
Netflix has outlined five key principles that define its approach to using generative AI responsibly:
- Generated materials must not replicate or recreate identifiable aspects of unowned or copyrighted content, and should not infringe on such works.
- The AI tools employed must not store, reuse, or train on production input or output data.
- Where feasible, these tools should operate within a secure environment to protect input data.
- Any generated material should be temporary and not included in final deliverables.
- Generative AI should not be used to replace or produce new performances by talent or covered union work without explicit consent.
Should partners feel confident in complying with these guidelines, they only need to notify their Netflix contact of their intent to utilize generative AI. However, if they are uncertain or believe they may not be in compliance, they are advised to seek clarification from their Netflix representative, as written approval might be necessary.
The posted guidelines emphasize the importance of navigating the legal risks associated with the use of generative AI. Netflix champions the notion that “audiences should be able to trust what they see and hear on screen,” asserting a commitment to preventing the technology from obscuring the boundaries between fact and fiction or inadvertently misleading viewers.
The announcement of these guidelines follows recent comments from Netflix co-CEO Ted Sarandos, who expressed his belief that AI has the potential to enhance the quality of films and series, rather than just reduce costs. He highlighted Netflix’s upcoming Argentinian sci-fi series The Eternaut as a case where generative AI might be applied to keep production expenses within budget. This indicates Netflix’s eagerness for creative teams to adopt such innovative workflows.
