Preventing misuse of AI voices. ElevenLabs to launch service compliant with C2PA standard, which provides tamper-proof digital signatures, in November
Note: the original Japanese article can be found at:
https://www.4gamer.net/games/991/G999101/20251029014/
Editorial Department: Yawaragi | 2025/10/29 03:39 (UTC)
ElevenLabs, a company specializing in AI voice synthesis software, has joined the international technical standard “C2PA (Coalition for Content Provenance and Authenticity)”, which verifies the origin and history of AI-generated content, and will begin offering services compliant with this standard.
C2PA is a technology standard established primarily by Adobe, Microsoft, and Google. It enables digital content—such as images, videos, and audio—to be embedded with tamper-proof digital signatures called “Content Credentials,” which record the creation and editing history of the content.
By complying with the C2PA standard, it becomes easy to identify and verify whether a generated voice was created by AI and where its origin lies, thereby establishing a foundational defense against misuse through deepfakes or impersonation.
Additionally, this helps reduce risks associated with unauthorized use of voice actors’ voices and protects rights while creating opportunities for monetization.
ElevenLabs states that it has already implemented independent technical safeguards to ensure high security even before adopting the C2PA standard. Specifically, it developed and operates a system that strictly verifies the match between the vocal fingerprint of provided audio data and the live voice of the individual when performing AI-based “voice cloning” technology—preventing cloning unless this verification is confirmed.
Furthermore, ElevenLabs embeds inaudible “digital watermarks” into voice content and uses its AI Speech Classifier to detect and identify these markers. This enables rapid tracing and identification of origins should any unauthorized use occur involving content generated through ElevenLabs.
Through the integration of these technologies with C2PA standards, it becomes possible to prevent illicit cloning at an early stage and enable provenance verification as well as post-event tracking—contributing significantly to enhanced transparency and safety across the industry.
The implementation of C2PA data embedding is scheduled to begin in November 2025, with all voice content generated through ElevenLabs receiving embedded Content Credentials.
Alongside this announcement, a statement from Yuki Kaji, the voice actor and producer behind the audio used in the vocal synthesis software “Soyogi Fractal,” has been released. Please refer to the official press release below for details.
ElevenLabs Official Website
＜The following is a direct reproduction of the manufacturer’s press release content＞
Dual Protection with VoiceCAPTCHA Technology and International Standard C2PA
ElevenLabs, committed to protecting Japanese creators’ voices and culture: Enhances its world-leading security measures with compliance to the international standard C2PA for ensuring trustworthiness in AI-generated audio.
As a leading company in AI voice synthesis, ElevenLabs (headquartered in New York State, USA; CEO: Mati Staniszewski, hereinafter “ElevenLabs”) has consistently implemented pioneering internal measures for AI governance and reliability assurance. To further enhance user confidence, we are now joining the international technical standard “C2PA (Coalition for Content Provenance and Authenticity)”, which verifies the origin and history of AI-generated content, and will begin offering services compliant with this standard.
Specifically, ElevenLabs has already implemented strict identity verification technology called VoiceCAPTCHA during registration for its professional voice cloning service (PVC). Additionally, we have rigorously applied three independent technologies: embedding proprietary digital watermarks into generated audio and using our AI Speech Classifier to detect and identify these markers—establishing world-class security measures ahead of industry trends. By combining this robust foundation with the global standard C2PA, we aim to deliver even greater peace of mind to users.
Strong Commitment from ElevenLabs to Address Japanese Challenges
ElevenLabs deeply respects Japan’s rich and diverse voice actor and content culture, exemplified by anime. We firmly oppose the growing number of incidents in Japan where voice actors’ or celebrities’ voices are used without consent for AI generation—so-called “unauthorized AI-generated voices.” We fully support the mission of the “NO MORE Unauthorized AI” movement and are committed to helping create a secure environment where Japanese creators and content industries can confidently use AI, through our strong technical framework combining our proprietary technology VoiceCAPTCHA with the international standard C2PA.
What is C2PA: An International Standard Ensuring Content Trustworthiness
C2PA is an industry-wide initiative established by Adobe, Microsoft, and Google. It defines a technical standard that embeds tamper-proof metadata—called Content Credentials—into digital content (images, videos, audio, etc.), recording the content’s creation and editing history (provenance). Compliance with C2PA enables users to easily identify and verify whether generated voice material was AI-created and trace its origin, forming a crucial foundation for preventing misuse via deepfakes or impersonation.
Specific Benefits of C2PA Compliance:
- Creators (Voice Actors & Performers): Reduces risk of unauthorized use of their voices and enables “provenance verification” for generated content—protecting rights and enabling monetization opportunities.
- Consumers & Users: Easily identifies whether the audio being listened to was AI-generated, thereby enhancing information reliability.
ElevenLabs’ Three Advanced Security Measures Combined with C2PA Standard
Since its founding, ElevenLabs has prioritized trustworthiness and security just as much as innovation. Prior to adopting the C2PA standard, we have consistently employed the following three independent technical safeguards:
Together, these measures aim to further strengthen responsible AI and lead industry-wide improvements in transparency and safety.
C2PA compliance adds the ability for global verification to our three existing protective features—further advancing industry-wide transparency and security leadership.
Comments from Voice Actor Yuki Kaji (Voice Actor & Producer of the “Soyogi Fractal” AI Voice Project)
Japan’s response to AI content lags far behind other countries. I feared that if this continues, we might face not only a decline in entertainment expression but even a loss of fundamental human dignity. While learning how to use AI aggressively, we must also prepare defenses—otherwise, one day it may be too late to reverse the damage. Yet, given the lack of comprehensive legal solutions so far, individual actions have clear limitations. I had nearly resigned myself to believing that a system truly capable of protecting Japanese voice rights would remain distant for years.
Then came this moment—the long-awaited arrival of VoiceCAPTCHA x C2PA, a tamper-proof vocal fingerprint watermark technology, now officially launched with full readiness.
This is nothing short of the “righteous protection system” we’ve been waiting for—truly trustworthy and reliable.
Whether you want to use your voice creatively, commercially, or simply for protection—the choice is yours.
Your voice rights belong only to you—not to anyone else.
The debate about AI’s merits has no bearing on this truth.
I sincerely hope that VoiceCAPTCHA x C2PA spreads rapidly across the world—eradicating harm caused by misuse of AI and shifting public perception toward a more positive view of it.
AI is merely technology, just a tool.
Whether we use it for crime or as a partner depends entirely on us.
The emergence of VoiceCAPTCHA x C2PA will undoubtedly mark a turning point in the history of AI.
Comments from ElevenLabs Founder and CEO Mati Staniszewski
Japan’s anime and voice acting culture is unparalleled worldwide, and we deeply admire its creativity. As an AI company, establishing ethics and trustworthiness is just as important a mission for us as innovation.
Our technology has always been secured under strict internal standards—but joining the C2PA standard offers us an opportunity to deepen understanding of our commitment to coexistence with Japan’s creative community. With unwavering determination to build a future where AI voice technology can be used confidently, we look forward to this initiative contributing to greater industry-wide transparency.
Technical Approach and Implementation Schedule for C2PA Integration
Implementation Timeline: Beginning in November 2025, gradual rollout of C2PA data embedding will commence.
Technical Approach: All audio content generated through ElevenLabs will embed Content Credentials (tamper-proof digital signatures)—proving the origin as ElevenLabs and enabling detection of malicious tampering.
About ElevenLabs
ElevenLabs, founded in 2022, is a global leader in AI audio research and technology. We develop cutting-edge AI audio tools for businesses, developers, creators, artists, and more. With a current market valuation of $6.6 billion (approximately ¥1 trillion as of October 2025), our platform serves over 40 million individuals and thousands of companies—including over 75% of Fortune 500 firms. Our services enable fast, large-scale creation of high-quality voiceovers at affordable prices, or launching interactive AI voice agents across more than 30 languages.