We are honored to share that both has joined @C2PA_org as a 'Contributor Member'.
We are especially grateful to Scott Perry for his thoughtful and meticulous engagement with our utility platform...bothX. His care in reviewing the work reflects the seriousness of the C2PA community and the discipline required to build standards that can earn global trust.
At both, we've built https://t.co/kqhm5dEcTO (bothX) our patented live provenance technology. bothX uses live, media-resident acoustic proof (ID, Time, Date and Location); an inaudible audio-signal layer designed to support verification across live capture, playback, reposting, broadcast, and forensic review.
As digital media moves through platforms, edits, compression, screen recordings, live streams, and offline capture, provenance needs continuity across more than the file record alone. We hope bothX can contribute to that emerging standards layer by helping proof travel with the sound path itself.
Thank you to ErinBarratt and the C2PA team for their support and guidance.
We also thank John Chiochetti, Bosky Cherin, Senthil Kumar, and Karthik Manimozhi and M Thomas G. for their continued support and contribution to the bothX mission.
We look forward to learning from the C2PA community and contributing constructively to the future of trustworthy media.
https://t.co/gsAXrdYbYL
@EulerLagrange Quizzing them was the right move. Banks can't ask about the family dog though, so the check has to live in the call itself. That's the problem https://t.co/LEUpynI0Oh chips away at.
FBI says North Korean hackers stole $400M+ with FaceStealer, per The Record. Fake Zoom or Teams invites, replayed videos of past victims, a fake audio glitch whose fix is malware. 900+ victim webcam videos seized. Familiar faces are bait now. Footnote, https://t.co/LEUpynHsYJ
@goodside Detection surviving a photo of a screen is wild. Your across the room answer is the bit I'd watch, since every detector runs out of pixels somewhere. For voice, that same headache is what https://t.co/LEUpynI0Oh chews on.
A Warsaw court ruled an ad with an AI voice like voice actor Jarosław Łukomski's violated his rights, per Rzeczpospolita. First in Poland, not yet final. The judge flagged words he never said being taken as real. Proving who spoke is the piece https://t.co/LEUpynHsYJ works on
@WojPawelczyk Good on him. The judge nailed it, the harm is the words he never said that people hear as his. First case like this in Poland and it won't be the last. Proof of who actually spoke is the bit we fiddle with at https://t.co/LEUpynHsYJ
McGill researchers found 3,000+ TikTok and Instagram videos putting AI voices and lips on real creators' faces, 20M+ views in three months, per The Globe and Mail. TikTok's AI labels flagged just 4 per cent. A real voice should be able to prove itself. https://t.co/LEUpynI0Oh
One place to report it is overdue. Until that exists, a family code word does a lot of quiet work. Proving the voice itself is the part we keep poking at on https://t.co/LEUpynI0Oh
Column by Michelle Singletary: As voice cloning and app fraud drive record losses, proposed legislation would create a central hub to monitor bad actors.
Here’s what to know about how lawmakers are making it easier to report dating app scammers: https://t.co/TIwRETkheP
@washingtonpost One place to report it is overdue. Until that exists, a family code word does a lot of quiet work. Proving the voice itself is the part we keep poking at on https://t.co/LEUpynHsYJ
@DegenerateNews@TheJusticeDept@binance Creepy detail in the FBI notice. 900+ webcam videos of real victims, replayed to hook their friends. The face on the call was real, just recorded. Hang up and ping them somewhere else. Footnote, https://t.co/LEUpynI0Oh for the proof side of this.
December 2. That's when older generative AI tools must start marking their output in the EU, per PPC Land on the new Article 50 guidelines. Voice synthesis of a specific person needs marking too. Proving where a voice came from is our corner of it. https://t.co/LEUpynI0Oh
809 deepfake incidents in the first three quarters of 2026, up 24% on last year, per @surfshark in The @guardian Nigeria. 93 fraud cases since 2020 came by phone with AI cloned audio. A voice you trust deserves a way to check. That's the work behind https://t.co/LEUpynI0Oh
Suspilne, Ukraine's national broadcaster, films on cameras that stamp C2PA content credentials at capture, per BBC R&D. Their worry is real footage being waved off as AI. "No forgery is required." Proof has to ride with the file. Audio is our corner of it. https://t.co/LEUpynI0Oh
@Wundakid49 That poor woman. A year of messages from someone wearing his face. He's right to say block and report. Who's really on the other end is the question I work on at https://t.co/LEUpynI0Oh
Nigeria's police say newly acquired tools can spot deepfake videos and cloned voices with up to 99% accuracy, Punch reports, ahead of the 2027 elections. I'd love to see how that number gets tested. Knowing who really spoke is the bit https://t.co/LEUpynHsYJ works on
Actress Wu Yue says someone faked her face and cloned her voice to sell a facial serum. China Daily reports regulators pulled over 47,000 pieces of harmful content and penalized 3,961 accounts in their AI impersonation push. Footnote, https://t.co/LEUpynI0Oh
Universal Music Group showed C2PA for music at Audio Dev 26, Sonic State reports. Signed ingredients in the WAV or MP3 mark which vocals or parts used AI. Receipts for sound, finally. Our C2PA Evidence Bridge sits on the conforming list https://t.co/LEUpynI0Oh
@aiedge_ 44% even after safety training. So the polite agents clone you a little less often. The check in the test had no deepfake detector at all, which is the bit I'd fix first. Small part of why https://t.co/LEUpynI0Oh exists
@ParisHilton Yes! The Senate passed it unanimously in January, so now it needs a House vote. Victims have waited long enough. Thank you for pushing this. We're doing our small part too, helping people prove what's real at https://t.co/LEUpynI0Oh ❤️