Microsoft Research Asia has developed VASA-1, an AI that creates realistic videos from a photo and voice clip.

VASA-1 can animate still images, such as the Mona Lisa, with lifelike lip-syncing and facial expressions.

The technology is designed for low-latency streaming, ideal for virtual interviews and video conferencing.

There are concerns about VASA-1's potential for misuse in spreading misinformation and ethical issues.

Microsoft sees VASA-1 as a tool for increasing educational access and aiding those with communication difficulties.

Other companies like OpenAI and Google are also exploring similar AI applications in art and media.

Microsoft will not publicly release VASA-1 until it can ensure responsible use and adherence to regulations.

The company is developing forgery detection methods to counteract the AI's potential for impersonation and fraud.