We've turned more than 150 real people into AI characters, from executives to event guests, and every one of them said yes first. Here's what deepfakes are, how to spot them, and how to make them the right way.
In November 2024, CBS Austin anchor Walt Maciborski ran a story on AI, and Steve did the demo.
Steve cloned Walt's voice from clips that were already public and laid it over his own. Then he put Walt's face over his own face. He also gave Walt some dance moves.
When Walt asked how people are supposed to tell the difference, Steve's answer was simple: "You can't." He said it comes down to media savvy and to asking where you're seeing something and who's behind it.
A deepfake is video, audio or an image made with AI to show a real person saying or doing something they never did. A few minutes of someone's voice or a handful of photos is now enough.
The same tools make digital doubles for films, event openers, training videos and translated content. The technology doesn't change between those uses. What changes is whether the person agreed to it.
We make these for a living, so here's the honest version. By eye, you often can't tell anymore.
There are still tells on cheaper fakes, like mouths that drift out of sync with the words, lighting on the face that doesn't match the room, hands and teeth that look wrong, and voices that are too smooth or oddly paced. The good ones don't have them.
The reliable test is the source. Where did the video first appear? Who posted it? Is a trusted outlet carrying it, or a site built to look like one? If something is designed to make you angry or scared fast, slow down before you share it.
We've made AI versions of executives at Hilton, Marriott, YPO, DTN and Texas Realtors for event films, plus hundreds of event guests through our video booths. Every one of them agreed first.
Our rule is written permission, every time. The person knows what we're making and where it will be used. When it matters to the audience, we say it's AI.
We won't make deceptive political deepfakes or non-consensual intimate images for any budget, and we won't make a likeness of anyone who hasn't agreed to it. Our full AI policy.
Checked 23 October 2026. This changes often, and it isn't legal advice.
Federal. The TAKE IT DOWN Act requires platforms to remove non-consensual intimate images, including AI-made ones, when they're reported. Those duties took effect in May 2026. The NO FAKES Act, which would create a federal right over your own face and voice, passed the Senate Judiciary Committee in June 2026 but isn't law yet.
Texas. In 2025, Texas expanded its deepfake law to cover non-consensual sexual images as well as video, effective September 1, 2025.
Actors. SAG-AFTRA contracts require consent and payment for digital replicas of union performers.
A deepfake is video, audio or an image made with AI to show a real person saying or doing something they never said or did. The same technology makes digital doubles for films and marketing. What separates the two is consent.
Often you can't, at least not by looking. The old tells, like odd blinking, mouths that don't match the words, strange lighting and mangled hands, are disappearing as the models improve. The better test is the source. Ask where the video came from, who posted it first, and whether a trusted outlet is carrying it.
Making a deepfake isn't illegal on its own, but many uses are. The federal TAKE IT DOWN Act requires platforms to remove non-consensual intimate deepfakes when notified, and those duties took effect in May 2026. Texas expanded its deepfake law in 2025 to cover non-consensual sexual images as well as video. The federal NO FAKES Act, which would create a right over your own likeness, passed a Senate committee in June 2026 but isn't law yet. This isn't legal advice.
Yes, with the CEO's written permission, and it's common for event openers, training and translated video. Get the consent in writing, say how the likeness will be used and for how long, and tell the audience it's AI when it matters.
It's a digital double made with the permission of the person it depicts. Studios call it a digital double, digital twin or AI avatar. The technology is the same as a deepfake. The permission is what makes it legitimate.
Only with that person's written permission. We won't make deceptive political deepfakes or non-consensual intimate images for any budget, and we won't make a likeness of anyone who hasn't agreed to it.
Steve Mudd, founder of Talentless AI, has demonstrated deepfakes on CBS Austin, has turned more than 150 real people into AI characters with their permission, and speaks on AI for Ragan Communications and the Austin AI Alliance.
Sources: CBS Austin, Skadden on the TAKE IT DOWN Act, Holland & Knight on the NO FAKES Act, Texas HB 449.
For brands, events and press. We're in Austin.
Let's talk