Amazon Releases Video Generator Only for Ads

Like Google, Amazon has released an AI-powered video generator, but it can only do a few things at a time and is only for advertisers

Amazon showed off a Video generator at its Accelerate conference today. This tool takes a single picture of a product and, after some processing, turns it into a video showcase of that product. The company says the tool can put together “custom” videos made by AI that “showcase a product’s features” for no extra charge.

One of Amazon’s blog posts says, “Videos from Video generator use Amazon’s unique retail insights to vividly bring a product story to life in ways relevant to customers.”

Jared Richman, VP of Amazon Ads, said in a statement that Video Generator will be improved over time before it is made available to more advertisers. Right now, Video Generator is in beta for some U.S. marketers.

Richman said, “Video generator is another important innovation that uses generative AI to spark creativity and give advertisers and shoppers more value.” “We are working hard to bring advertisers generative AI apps that let them make ads that look great and get great results.”

Live image, a related new feature revealed today, turns a still frame into an animated GIF that is only a few seconds long. It’s also in limited beta and is part of Amazon’s AI-powered image generation suite, Image Generator. Amazon didn’t show any sample footage from Video Generator and only gave a few technical details about the tool and live pictures. Some things aren’t clear, like how long generated clips can be or what the highest quality can be.

Our team has asked Amazon for more details and will change this post if we hear back.

Amazon’s move into generative video comes simultaneously with other companies releasing video-making tools. This month, startups Runway and Luma released generative video APIs. Google also said it is starting to add its primary video model, Veo, to YouTube Shorts. In other news, Adobe noted that video creation would be added to Creative Suite by the end of the year, and OpenAI is set to release some version of its technology, Sora, this fall.

When you use these tools, there are risks, just like with any other creative AI system.

Video-making programs are taught by showing them a vast number of videos so that they can “learn” how to make new clips. Some sellers train models on copyrighted videos without getting permission from the owners or makers of the videos. When these models “regurgitate” copyrighted stills, people who use them are at risk of IP lawsuits.

As part of its compensation policy, Amazon is one of several companies that have said they will defend customers accused of copyright violations involving media made by its models. We asked the business if that policy covers video generators and live pictures.

No matter how the court cases turn out, one thing is becoming clear: generative AI could change the movie and TV business as we know it. The Animation Guild, a group for Hollywood animators and cartoonists, paid for a study in 2024 that says generative AI will take away more than 100,000 entertainment jobs in the U.S. by 2026.