Deepfakes are rising — and so is the risk to truth. At UM-Flint, researchers at the SMILES Lab built DeepTect, an AI tool that detects and explains deepfakes before the damage is done. From courtroom evidence to cloned voices, DeepTect helps expose what’s fake — clearly, quickly, and credibly. Watch this video from Michigan Research as AI expert Khalid Malik explains.

Because of partnerships with federal agencies like the National Institutes of Health and the Department of Defense, Karl Jepsen has worked for decades to build up the field of bone research. “We are just now seeing the outcomes of funded projects from 20, 30, 40 years ago,” he says.

The ZEUS laser facility has roughly doubled the peak power of any other laser in the U.S. with its first official experiment at 2 petawatts (2 quadrillion watts). “This milestone marks the beginning of experiments that move into unexplored territory for American high field science,” says Karl Krushelnick, director of U-M’s Gérard Mourou Center for Ultrafast Optical Science.

Marcus Foundation’s $30 million gift supports a collaboration between Stanford and U-M to help stroke victims regain the ability to read, write and speak. The scientists are using the world’s smallest computers linked up to the world’s most biocompatible sensors.

Desalination plants, a major and growing source of freshwater in dry regions, could produce less harmful waste using electricity and new membranes made at the University of Michigan.

A new class of semiconductors that can store information in electric fields could enable computers that run on less power, sensors with quantum precision, and the conversion of signals between electrical, optical and acoustic forms. But how they maintained two opposite electric polarizations in the same material was a mystery. Until now.

Using NASA’s James Webb Space Telescope, a team of researchers, including astronomers from the University of Michigan, are closing in on the answer to a looming cosmic question. In probing the Flame Nebula, they’re finding out what’s the smallest celestial body that can form on its own from clouds of gas and dust in space.

The project assembles a team of experts from across the country for a mission called STARI — STarlight Acquisition and Reflection toward Interferometry. The goal is to showcase the viability of a new technique for studying exoplanets, or planets outside of our solar system.

It’s easy to think that machine learning is a completely digital phenomenon, made possible by computers and algorithms that can mimic brain-like behaviors. But the first machines were analog and now, a small but growing body of research is showing that mechanical systems are capable of learning, too, say physicists at U-M.