Microsoft has filed a patent for an AI system that can modify live speech in video, including redrawing lips to match. The filing US 2026/0341309 A1 was submitted in April 2025 and published October 29, 2026. An AI agent scans live audio and swaps or removes words based on rules and viewer profiles like age, language, or location.
A generative model replaces the words in both audio and video with a one-to-ten-second delay, running locally on AR glasses, phones, tablets, or cameras. Filler words, stutters, mispronounced words, and colloquial terms like 'y'all' or 'bite the bullet' are listed as replaceable items, with new words sourced from internet searches. The filing covers lectures, award shows, game shows, and live sports.
Sony is also active in AI patents.
Source: TechPowerUp News · Summarized by HeadlinesBriefing