This Variety Intelligence Platform report examines the current use and impact of generative AI in the US film and television industry. The report is based on interviews with industry experts and surveys of managers, employees and consumers in the media and entertainment sector.

Introduction

Generative AI is increasingly being used in Hollywood, but the way, how it is used, is not transparent. There are different opinions on the use of AI, Some are in favour of the technology's potential, while others are cautious due to restrictions and risks. One of the most important concerns is the question, whether the quality of the AI tools meets the high production standards, as well as legal and ethical issues relating to copyright and intellectual property.

Current status of the introduction

  • Previsualisation: AI is used in conceptual art, to create initial reference materials and speed up the coordination process between the parties involved. However, there are concerns about copyright and the quality of the AI-generated images.
  • Production and post-production: Video generation models such as Sora from OpenAI and Veo from Google DeepMind are making progress, however, are not yet suitable for large-scale use in premium content due to quality and controllability problems. Deepfake technologies are used for minor cosmetic improvements, face exchange and lip synchronisation.
  • Localisation of content: AI synchronisation is used for content with lower requirements, as the quality cannot yet keep up with traditional dubbing. Lip-synchronisation with AI is being tested by major Hollywood studios, to determine, whether it increases audience reach and engagement.

Ethical implementation

The introduction of generative AI in Hollywood is being hampered by concerns about copyright infringement, performance and ethical issues. There are uncertainties about copyright laws in connection with AI-generated content and the use of copyrighted data for the training of AI models. There is a growing need for ethical AI training methods, that use less harmful data and ensure the consent and compensation of authors.

Surveys on generative AI and entertainment

  • Introduction: The majority of decision-makers in the media and entertainment (M&E) sector state that that your company is researching generative AI, tests or uses.
  • Advantages: Decision-makers assume this, that generative AI increases productivity, will improve the quality of work and shorten project lead times.
  • Implementation: The most important factors, that influence the introduction of generative AI, include the quality of the AI-generated content, the effectiveness of the AI tools, data protection, cyber security, the copyright, ethical data sources, acceptance by consumers and acceptance by employees.
  • Effects: It is expected, that generative AI will have a major or minor impact on most creative roles in the entertainment industry, whereby animators, VFX artists and dubbing actors will be the most affected.
  • Ethics: Both employees in the entertainment sector and consumers agree, that consent should be required for the use of AI to create replicas of style or appearance.

Conclusion

The introduction of generative AI in the film and television industry is underway, but also harbours challenges and risks. The success of the launch depends on the further development of AI technology, clarification of legal issues and the development of an ethical framework for the use of AI.

 

If you are interested in current developments in AI from the perspective of the American trade press, it is worth purchasing the full 33-page report overall!

Graphics (some excerpts)

Adoption of generative AI in US media and entertainment companies

Assumption of generative AIPercentage of decision-makers in the media and entertainment sector
The company has currently banned its use6%
We have not yet begun to explore the possibilities15%
 

Expected impact of artificial intelligence on TV touchpoints, according to industry experts

RangeMajor effectsModerate impactLittle to no impact
Content recognition (search and recommendation for users)81%17%2%
    
Content creation (scriptwriting, voiceover, animations)55%37%8%
   
Video editing and post-production46%49%5%