Respeecher ads

A lot of my job is therapy. Only half-joking. We talk a lot about fear that always comes from the unknown. Talents are afraid of new technology and community reaction. Studios are afraid of unions and audience response. Don't get me wrong - we have excellent technology behind Respeecher. But without trust from the industry, we wouldn't get where we are now. Working in constant AI fear, we do a lot of work to educate people - how the tech works, what possibilities it brings, what it's best suitable for. What I think move it past fear: 1/ Regulations for the space 2/ Knowledge about what this tech is and what it is not - differentiate AI tools 3/ Transparency and control between actors, studios, and audiences But none of it happens by itself. Be honest: when you hear “AI in film/games/your field,” what’s the first thought that comes to mind?

Over 2,000 verified deepfake attacks were logged in just one quarter of 2025. I wish that was a typo. The least any of us can do is know enough so we don't sign the wrong clause, end up in the wrong headline, or discover our voice was used in a way we never agreed to. I’m sure any decision about AI use can’t be made without a clear set of principles. If only there was a shared checklist of “this is okay” and “this is not”… Oh, here it is. Our team composed this cheat sheet featuring the same framework we use ourselves - the things everyone should have on the table when deciding whether to work with AI voice at all. It’s just a snapshot of the complete guide that is honest, practical, free. Useful if you work with AI every day, and just as useful if you don't but want to understand what's happening around you. Link in the first comment!

Martin Scorsese advising an AI startup might be the best‑case scenario for the industry Black Forest Labs, makers of FLUX image models, just announced that Scorsese’s joining as an advisor. He’s already using their AI to storyboard, saying it helps him show what he sees in his head more clearly. I wanted to see how this lands with people. Reddit’s updating at about a comment a minute, and reactions mostly fall into three camps 1/ "Even him" The patron saint of cinema went to an AI startup with roots in the early messy days of model training and copyright debates 2/ “It's just a tool” He's using it for storyboards only. Pre-production communication - totally fine 3/ “This is how power works” (the one I find most honest) He gets a carefully managed tool he fully controls - everyone else gets the downstream version of that decision In the end, this is all about who gets the tool, and who becomes the input. If people like Scorsese use their seat at the table to push on consent and training data, that’s the best version of this story. I hope that’s what happens. So before we jump to “hypocrite”, I’d rather ask: Are we actually angry at the person, who’s spent a lifetime in film industry, or at everything generative AI has already broken in there?

I don't blame anyone who's suspicious of AI voice. When you’ve seen how casual consent or usage limits are treated sometimes, the skepticism is pretty rational. But I've also seen that when you follow a few simple rules, AI doesn't have to be scary. It’s mostly common sense - ask before you train, write down what's allowed, label what's synthetic, never skip human oversight. That’s what this cheat sheet is - one page on how to stay on the right side of the law, plus the questions to push back on any AI vendor. It’s for sound supervisors, audio leads at studios, producers and heads of post. Or just for those who are skeptical about AI or just tired of hand‑wavy “ethical” talk. All the info below comes from a 15‑page guide we did on synthetic media, with a lot more detail, examples, and things you should know. And it’s free. I’ll leave a PDF link in the comments.

Heard basement superhero movies are around the corner. Will they have basement‑level sound too? Only a year ago people were hesitant to admit they used AI in production. This year at the Cannes film festival you could almost feel it in the air - “AI is here, fighting it is a losing battle”. The thing is, I don’t think that was ever the right battle. The real one starts after “yes” - which tools, trained on what, with whose permission, and would you trust that voice next to what you recorded on set. Majority didn’t discover AI through policy docs. They met it in headlines, where voices cloned and faces faked without consent. Now even when a project does it properly, “AI voice” in credits reads as a stolen one. And that's not on them - every shortcut reinforced that. So no, accepting AI was never the hard part. Earning back the audience after broken trust is. I wonder, when you see AI in the credits, is that already a red flag for you?

Look what you made her do: Taylor Swift has trademarked her voice to fight AI deepfakes. It centers on three U.S. trademark filings: sound marks for “Hey, it’s Taylor Swift” and “Hey, it’s Taylor,” and a visual mark for one pink‑guitar Eras Tour photo. Together they frame her voice and likeness as enforceable IP against AI uses she never approved. A true mastermind move: today’s copyright rules protect finished works – tracks, lyrics, recordings. They don’t cover the way a person’s voice sounds. This “trademark yourself” strategy gives celebrities an extra federal lever (alongside publicity and privacy law) to challenge confusingly similar AI imitations and demand removal. And we’re enchanted to see this! 💜