Lionsgate’s new partnership with Runway, the video AI startup, is one of the most formal embraces of AI technology yet seen by an entertainment company. However, while the deal may cause excitement among brands that want to create more content based on their intellectual property, it may deepen fears among creatives that AI is deflating their value.

This creative-AI tension has been building over the past year as the advertising industry has expanded its use of AI tools in ad production. Brands such as Under Armour and Toys R Us have released ads developed with generative AI, showcasing the technology’s ability to deliver a range of creative outputs in a short amount of time. But with each instance of AI-powered production, human creatives have voiced protest. Research shows that consumers feel similarly, with fewer than half wanting brands to use AI to create ads. 

Nonetheless, brands continue to be enticed by the technology, and Lionsgate’s plans help to illustrate why. At the heart of its partnership is a plan for Runway to train a new model based on the studio’s portfolio of film and TV content. Lionsgate will use this model to generate cinematic video, beginning with storyboarding before shifting to actual film content, such as background generation and special effects, according to The Wall Street Journal, which first reported on the partnership.

Numerous brands have fine-tuned AI models using their proprietary data to generate text, images and videos that are more relevant to their businesses. However, rarely have any marketers developed entirely bespoke models based on their full library of information. Doing so could allow Lionsgate to produce a large volume of content at a fraction of the traditional cost, said Jay Pattisall, VP and principal analyst at research firm Forrester.

Lionsgate’s example could be an exciting prospect for consumer brands eager to turn a pool of IP into usable media, whether for casual consumption on social platforms or longer-form entertainment. Runway is just one of a handful of AI video products developing this opportunity for marketers. Showrunner, for example, is a tool that aims to generate full-length TV shows built from pre-existing IP. Showrunner’s early prototypes have uncannily imitated Disney’s “Cars” franchise and “South Park.”

Tech giants are wading into similar waters with video AI. Last week, Amazon rolled out a video generator for advertisers that can take an image of a product and turn it into a short clip and Google announced that it is integrating video generation capabilities into YouTube Shorts.

What remains unclear about any of these AI-produced videos, including those born out of Lionsgate’s model, is whether they will be ready for mass consumption. It is unlikely that Lionsgate will go to market anytime soon with long-form productions generated with the help of Runway, Pattisall said.

A Runway spokesperson declined to comment for this story, instead pointing to a quote from a blog post announcing the partnership.

It could be years before Lionsgate puts AI-generated long-form video in front of audiences, according to Champ Bennett, co-founder and CEO of Capsule, an AI video editing platform. Bennett said that while AI’s abilities to produce video have greatly improved within the past few months, this content is nowhere near the level of quality it needs to be for primetime, especially for a Hollywood studio.

Bennett voiced particular concern about how AI can be a nuisance in the production process. For instance, directors typically maintain total control of shots in order to capture the best scene. But with the current models, “you get about two percent of that control,” Bennett said, as the AI wields cinematographic authority.

Recent news: 9 AI tools agencies are embracing right now

AI’s shortcomings mean that Lionsgate will most likely use Runway to make minor adjustments to non-AI-generated film, such as changing a character’s eye color, Bennett said. The studio’s initial emphasis on using AI in storyboarding supports this point, too. Limiting the scope of AI’s work is similar to the hedging brands are doing when working with AI. Many in the ad industry see AI staying in the background of the creativity process, augmenting ideation, perhaps, but never actually becoming capable of producing content for mainstream audiences.

Of course, others disagree. AI-generated video is nearly ready for mass audiences, according to Matty Beckerman, a former film producer and founder and CEO of advertising technology company IRCode.

“We are moments away from something being of high enough quality,” Beckerman said. 

When that time comes, the floodgates will open for other film studios to follow suit, Beckerman added.

Tension deepens

Against the backdrop of Lionsgate’s new partnership, fake ads that look real routinely make the rounds on social media, further intriguing brands while creatives are left to wonder about their future value to the production process. 

An anti-AI movement is gaining steam in response to this tension, which was evident on social media in the wake of the announcement of Runway and Lionsgate’s partnership.

Lionsgate will continue to face scrutiny about this deal, especially after it reveals the AI model’s first fruits, Forrester’s Pattisall said. Brands are familiar with this kind of vigilance. Consumers have an eye for AI-generated work—even when it’s not actually there—and are quick to raise alarm if it feels out of place.

More: AI in ads divides consumers—what brands should know

Recognizing the power of angry consumers, some companies have vowed to keep AI in its place. Last month, design platform Procreate announced it would not integrate the technology into its products, which triggered a surge of support from the creative community. This group simultaneously took shots at Adobe, which directly competes with Procreate and has widely embraced AI, for not protecting the industry. Brands including Dove have issued restrictions on the use of AI following public flops made by the technology.

Other media brands, too, will be watching closely to see how Lionsgate’s partnership shakes out, Bennett said. Most companies are likely having conversations around how to implement AI, but few have yet to declare their efforts in such a public and concentrated way as Lionsgate.

For now, Lionsgate and other media brands will need to contend with copyright issues facing the production of AI-generated content. The U.S. Copyright Office has said that copyright protections will not be afforded to material mainly created by AI. However, the USCO made this position a year and a half ago, when AI models looked quite different. It’s unclear how a model that is trained entirely from a company’s library of content, as in the case of Lionsgate’s partnership with Runway, fits into the USCO’s framework.

“This [deal] is a bet, but it’s not a sure thing,” Bennett said.