Real-Time AI Video Generation Arrives, Raising Concerns

New technology from Runway and Nvidia can produce video in under 100 milliseconds, blurring the line between real and fake.

Mar. 21, 2026 at 5:30pm

Got story updates? Submit your updates here. ›

A new real-time video AI model was demonstrated at Nvidia's GPU Technology Conference, capable of generating its first frame in less than a tenth of a second. While this technology could lead to advancements in virtual reality and interactive experiences, it also raises concerns about the potential for misuse, with AI-generated content becoming increasingly difficult to distinguish from reality.

Why it matters

The rapid development of real-time video AI generation technology has significant implications for the future of media, entertainment, and even politics. As the ability to create convincing fake videos in an instant becomes more accessible, it could enable the spread of misinformation, manipulation, and deception on a larger scale, challenging our ability to discern truth from fiction.

The details

The new video model, developed by Runway in partnership with Nvidia, can generate high-definition video in under 100 milliseconds, which is faster than the blink of an eye. This represents a significant advancement in AI-powered video generation, as previous models often took several seconds or longer to produce a single frame. The demonstration at Nvidia's GTC event showcased the potential for this technology to enable real-time, interactive virtual experiences, but also highlighted the risks of such capabilities falling into the wrong hands.

  • The new real-time video AI model was demonstrated at Nvidia's GPU Technology Conference in San Jose on March 20, 2026.

The players

Runway

A company working on developing AI-powered video generation and interactive virtual experiences.

Nvidia

A technology company that specializes in graphics processing units (GPUs) and AI-focused hardware, including the Vera Rubin supercomputer used in the demonstration.

Vera Rubin

An AI-focused supercomputer developed by Nvidia, featuring 36 Vero CPUs, 72 Rubin GPUs, 54 terabytes of CPU memory, and 20.7 terabytes of GPU memory, capable of running the real-time video AI model.

Got photos? Submit your photos here. ›

What they’re saying

“Well, on the good side, it's definitely a step towards the holodeck-style convergence I've been banging on about for years now, in which we'll be able to interact in real time with characters, situations and worlds of our own choosing. The ultimate video game experience in virtual reality, delivered frame by frame, à la carte, in response to your every whim.”

— Freedomsphoenix Readerfour, Author

“On the bad side ... Well, everything that's now taking you a second glance to spot as AI will at some point start coming at you in real time, tailored to convince or persuade you, generated using real images as a starting point, and potentially capable of reading and responding to your body language in real time. Such a strange world we're constructing for ourselves!”

— Freedomsphoenix Readerfour, Author

What’s next

As this technology continues to advance, it will be crucial for governments, corporations, and the public to address the potential risks and develop strategies to mitigate the misuse of real-time AI-generated video content.

The takeaway

The arrival of real-time AI video generation technology represents a significant milestone in the evolution of media and entertainment, but it also raises serious concerns about the potential for deception, manipulation, and the erosion of trust in digital content. As this capability becomes more accessible, it will be essential for society to find ways to adapt and ensure the responsible development and use of this powerful technology.