PULSE the living trend engine
🤖 Open Intelligence Dossier available for AI agents & citation View Markdown (.md) →
◼ Archived Technology

Microsoft Patents an AI That Can Literally Put Words in People’s Mouths in Real Time

Microsoft has patented an AI technology capable of altering speech and lip movements in real time during live video broadcasts.

5sources
5articles
3velocity
+40%since first seen
15d agofirst detected

Velocity

How fast coverage is spreading — measured hourly from article rate × source diversity. How this works →

The brief

Microsoft has filed a patent for a new artificial intelligence system designed to modify what individuals say during live video transmissions. According to reports from Digit and Futurism, this technology is capable of removing unwanted words and fixing speech errors as they happen. The system does not only alter the audio output but also modifies the visual representation of the speaker. As detailed by WION and India Today, the AI can change a person's words and then reshape their lips and face to match the new audio in real time, ensuring that the visual movements align perfectly with the altered speech. Coverage of this development is widespread across several tech and news outlets. Digit emphasizes the utility of the tool for live broadcasts, focusing on its ability to eliminate errors during a stream.

WION and Tech Times specifically highlight the synchronization aspect, noting that the AI shapes lips to match the changed words. Futurism characterizes the capability as putting words in people's mouths in real time. These reports collectively describe a system that integrates audio manipulation with sophisticated visual deep-learning techniques to maintain a seamless appearance during live transmissions. The significance of this patent lies in its application to live media and communication. Until now, most AI-driven speech and video alterations required post-production processing. The ability to implement these changes instantaneously during a live TV broadcast or video call represents a shift in how digital media is consumed and produced.

By allowing a broadcaster or user to remove mistakes or change phrasing on the fly, the technology introduces a layer of real-time curation to human speech that was previously impossible in a live setting. Future developments will likely center on the deployment and practical application of this patented technology. While the current coverage confirms the existence of the patent and its technical capabilities, it does not specify a release date for a consumer product. Observers will be watching to see if Microsoft integrates this tool into existing broadcasting software or communication platforms. The focus remains on how the ability to change words and faces in live videos will be managed and whether the company will provide disclosures when the AI is active during a broadcast.

Synthesized by PULSE from the headlines below under a strict no-invention contract. ✓ fact-checked: all claims supported by sources Updated 14d ago.

Quick answers

What does the patented Microsoft AI do?

It can remove unwanted words and fix speech errors in real time during live broadcasts, while simultaneously adjusting the speaker's lips and face to match the new words.

Which outlets reported on this patent?

The technology was reported on by Digit, WION, India Today, Tech Times, and Futurism.

Does the AI only change the audio?

No, according to WION and India Today, the AI also changes the speaker's face and shapes their lips to match the altered speech.

Coverage (5)

Topics

Related trends

\n \n \n \n \n \n \n