Meta has unveiled a new AI-powered video generator, Movie Gen, which produces high-definition video clips with sound through simple text prompts. This marks a significant step for Meta, coming shortly after OpenAI introduced a similar model, Sora, earlier this year. However, unlike OpenAI’s model, Meta’s Movie Gen isn’t yet available for public use.
Movie Gen is designed to generate new video content based on textual input and can also edit existing footage or images. The New York Times reports that the tool automatically generates accompanying audio to align with the visuals, adding sound effects, ambient noise, and background music to enhance realism. Users can specify different aspect ratios for the videos, making the tool flexible for various media formats.
Beyond creating entirely new videos, Movie Gen offers features for customizing visuals from still images or modifying elements within existing footage. One demonstration from Meta shows how a simple headshot of a woman can be transformed into a lively video where she is seen relaxing in a pumpkin patch with a drink. This highlights the tool’s capacity to take static images and give them dynamic life through generated scenes.
Additionally, Movie Gen allows for creative video edits by adjusting the style, transitions, or content of pre-existing footage. Meta provided an example of an animated runner in a video, which Movie Gen altered in several ways: in one instance, the runner held pompoms; in another, the backdrop shifted to a desert; and in a third, the character donned a dinosaur costume. Such changes can be applied simply by inputting text commands, showcasing how Meta’s model can use AI to transform video elements flexibly and creatively.
The announcement comes nearly two years after the release of popular AI image and video generators, signaling another leap forward in multimedia technology. Over the past six months, major tech companies, including Google and OpenAI, have been developing similar video tools. OpenAI’s Sora, which was revealed in February, is still in development and has not been released to the public. Notably, one of Sora’s co-leads recently left OpenAI to join Google, signaling increased competition and talent mobility in the AI field.
Despite the advances in video generation technology, Meta’s chief product officer, Chris Cox, indicated on Threads that Movie Gen’s public release is still some way off. He mentioned that the tool remains costly to operate and requires significant time for each video to generate, which are factors delaying a wide rollout. The demands of this technology highlight the challenges in scaling AI-powered video generation for broad usage.
As with AI image generators, AI video generators raise questions about ethics, ownership, and the potential for misuse. Concerns have already emerged about how these tools might violate copyright policies. Reports suggest that Runway, a prominent AI startup, trained its video model on YouTube content without authorization, an approach that YouTube’s CEO, Neal Mohan, confirmed would breach the platform’s policies. Meta, for its part, has stated that Movie Gen was trained on “a combination of licensed and publicly available datasets,” but the company has not specified which sources it used.
The development of AI video generators has sparked broader debates among creative professionals, including filmmakers, writers, actors, and visual artists, who worry about how this technology could disrupt their industries. AI’s growing influence has been a focal point in several high-profile labor disputes, including last year’s joint Hollywood strikes by the Screen Actors Guild – American Federation of Television and Radio Artists (SAG-AFTRA) and the Writers Guild of America (WGA). These unions and their members fear that the rise of AI-generated content could lead to reduced demand for human creators and potentially diminish their income and creative control.
In addition to concerns within creative industries, the broader use of AI video generation raises questions around the authenticity of media. With the ability to generate or modify video with such precision, there are worries about potential misinformation and deceptive uses.
Video editing software already enables significant manipulation, but AI tools like Movie Gen offer a level of ease and sophistication that could make altering reality much more accessible. As a result, some experts argue that there needs to be stricter regulation or clearer guidelines around AI-generated video content to protect against misuse.
Meta’s introduction of Movie Gen underscores the rapid evolution of AI in the multimedia space. As companies race to bring similar tools to market, questions of accessibility, ethics, and impact on creative industries are likely to remain central to discussions around these technologies.
The coming years will determine how AI video generation is integrated into both commercial applications and creative work, as well as how it is received by the public, creatives, and regulators. For now, Meta and its competitors continue to push the boundaries, setting the stage for a new wave of AI-driven multimedia content that could reshape how we experience and create visual media.