Microsoft Patents Live Streaming AI

Microsoft has recently submitted a patent for an artificial intelligence system capable of modifying what individuals say during live video broadcasts. The proposed alteration would occur within a brief timeframe, ranging from one to ten seconds, and would involve both the audio and video elements to ensure the change remains undetected.

The patent, titled "Near Real-Time AI Enhancement of Video or AR Presentations of Live Events" (US application 20260341309), suggests the system would replace certain words and subsequently redraw the speaker's mouth to align with the new spoken text. According to the filing, the AI has the potential to eliminate filler words like "um," slang, mumbled or mispronounced words, and even offensive language.

Furthermore, the replacements could be customized based on the viewer's profile, encompassing factors such as age, location, and language. Another claim within the patent pertains to live translation accompanied by lip-sync. The potential implications of this technology are considerable, as the editing occurs in real-time during the video stream, meaning viewers would be unaware of the actual spoken content.

This presents a risk of spreading misinformation and engaging in fraudulent activities. It is worth noting that the filing contains just one dependent claim that preserves the original recording. As it stands, this is a patent application and not an actual product, and Microsoft has not announced any plans to release the technology.

Microsoft has been increasingly investing in artificial intelligence, with recent developments including a significant overhaul of its Copilot feature and the creation of an AI system designed to diagnose diseases.