Microsoft's Asia research team has developed VASA-1, a highly advanced AI model capable of producing hyper-realistic talking face videos from static images and audio.

The VASA software uses machine learning to perfectly sync lip movements with audio and convincingly simulates facial expressions and head movements.

There is a rising concern over the potential misuse of such technology, including spreading false information and conducting scams.

Microsoft is withholding the release of the technology to the public, aiming to ensure its responsible use and adherence to forthcoming regulations.

The European Union is actively considering implementing regulations for generative AI to address these concerns about its misuse.