Amazon unveils AI models to rival Adobe, Meta, OpenAI and more; see details

At its annual AWS conference in Las Vegas, Amazon unveiled a new suite of artificial intelligence (AI) platforms, called foundation models, capable of generating text, images, and videos. Dubbed “Nova,” these tools represent Amazon’s most ambitious step yet to establish itself as a leader in generative AI, challenging competitors like Adobe, Meta, and OpenAI.

Amazon CEO Andy Jassy introduced the Nova models during the event on Tuesday, emphasising the company’s focus on meeting developer demands. “They want better latency. They want lower cost. They want the ability to do fine-tuning,” Jassy said, highlighting the priorities driving the innovations.

Rohit Prasad, Amazon’s head of artificial general intelligence, said the new models would stand out due to their speed and cost-efficiency. “If I have something better to offer, then customers will come and use it,” Prasad said in an interview. While acknowledging the industry’s early stage, Prasad added that Amazon has the opportunity to take the lead as AI technology evolves.

One of the standout offerings is Nova Reel, Amazon’s new software for video generation. Users can create six-second videos from a single image or text prompt, with plans to support videos up to two minutes in the near future. These tools aim to revolutionise product displays on platforms like Amazon’s marketplace, while also appealing to the entertainment industry for faster and more efficient filmmaking.

However, the rising popularity of such tools has sparked concerns over potential copyright infringements. Amazon responded by committing to watermarking and other measures to prevent the misuse of its AI tools.

Another major launch is Canvas, a platform that generates images from short text prompts. Amazon also announced plans to release a multi-modal AI model capable of processing and generating text, images, speech, and video. This marks a significant leap in AI’s ability to create integrated solutions for industries ranging from e-commerce to entertainment.

In addition, the company plans to integrate its revamped AI into Alexa, Amazon’s popular voice assistant. Internally known as “Banyan,” the project has faced delays over concerns about the system’s accuracy and responsiveness. While Prasad declined to provide detailed updates, he expressed optimism about the future of voice assistants.

The announcements come as Amazon works to shake off perceptions that it was lagging behind rivals like Adobe and Meta in developing advanced AI applications. With these new tools, Amazon hopes to capture market share in video and image generation—a hotly contested field with major players like ByteDance and OpenAI already in the game.

Amazon’s push into AI follows recent efforts by competitors to deploy similar technologies. Meta and Adobe have recently introduced tools for generating videos and images from text prompts, while OpenAI has continued to refine its text and image generation capabilities.

Jassy stressed the importance of responsible AI use during his keynote, highlighting Amazon’s commitment to watermarking and ethical safeguards. “We want to ensure that these tools are used responsibly to prevent the spread of harmful content,” he said.