How Are Global Laws Responding to AI Voice Cloning Without Consent?

Lithuania's Proposed Voice Protection Bill

Lithuania has introduced a bill that would create a new legal category for voice as a protected personal attribute, making unauthorized AI cloning a civil and criminal offense. The proposal would require explicit consent from the voice owner before any recording, processing, or synthesis for commercial or public use, and it would shift the burden of proof onto the platform or user who deploys the cloned voice. If enacted, Lithuania would become one of the first European states to treat voice itself—rather than the underlying recording—as an intangible property right, giving individuals the power to sue for damages and demand takedowns without relying on copyright or personality-rights doctrines that were designed for an analog era.

Also worth reading: What Are the Best AI Voice Cloning Tools for AI Voice Actors? · What Should AI Voice Cloning Contract Clauses Protect for Performers? · Can AI Voice Actors Clone Your Voice Without Permission?

Across the globe, lawmakers are scrambling to close similar gaps. In the United Kingdom, the Data Protection and Digital Information Bill proposes that synthetic voices be labeled as “synthetic media,” with penalties for misleading use, while the proposed AI Act in the European Parliament would classify real-time voice replication as a high-risk application requiring conformity assessments. Australia’s Copyright Inquiry is examining whether a performer’s voice should receive neighboring-rights protection akin to sound recordings, and Mexico’s lower house has passed a measure obligating social-media platforms to remove AI-altered voices within twenty-four hours of a takedown notice. The United States, by contrast, remains fragmented: some states extend right-of-publicity statutes to voiceprints, but federal legislation is still in draft form. Together, these initiatives signal a rapid shift from treating voice cloning as a novelty to regulating it as a fundamental threat to identity and trust.

UK Legal Gaps in Voice Rights

Global lawmakers are scrambling to catch up with the rapid rise of AI voice cloning, and the UK’s current framework is lagging behind several international efforts. While the UK has no specific legislation addressing non-consensual voice replication, other jurisdictions are moving decisively. Lithuania has proposed a bill that would grant individuals legal safeguards against unauthorized use of their voice, treating it as a personal right akin to image protection. Meanwhile, Australia is exploring whether existing copyright law can extend to vocal performances, with scholars arguing that voice should be recognized as a distinct intellectual property form. These developments reflect a growing consensus that voice is not just a biometric trait but a legally protectable expression of identity.

In parallel, Mexico has taken a platform-centric approach, voting to require internet companies to remove AI-altered voices upon request, shifting responsibility to tech intermediaries. This contrasts with the UK’s more passive stance, which relies on general data protection and fraud laws rather than voice-specific statutes. As AI voice actors become indistinguishable from real humans, the absence of targeted regulation in the UK leaves individuals vulnerable to impersonation, deepfake audio, and commercial exploitation without clear legal recourse. The international trend suggests that voice rights are emerging as a distinct legal domain, and the UK risks falling behind in protecting its citizens from digital voice theft.

Australia's Copyright Approach to Voices

Global legal systems are scrambling to keep pace with the rapid advance of AI voice cloning, revealing a patchwork of responses that often lag behind the technology. Lithuania has emerged as a proactive force, with lawmakers proposing dedicated legal safeguards that would explicitly protect individuals from having their voices cloned without consent. This initiative reflects a growing recognition that voice, much like image or likeness, requires distinct legal recognition in the digital age. Meanwhile, the United Kingdom faces scrutiny as existing laws may prove inadequate to address the nuances of voice cloning, leaving citizens potentially vulnerable to exploitation. The BBC's investigation highlights how current frameworks struggle to define voice as property or protect it as intellectual creation, creating legal gray areas that bad actors could exploit.

Mexico has taken a more platform-centric approach, voting to require digital platforms to remove AI-altered voices that violate rights, shifting responsibility to intermediaries. This mirrors broader trends in content moderation but raises questions about enforcement and jurisdiction. Australia's copyright framework, as analyzed by Wolters Kluwer, offers a potential model by treating voice as a protectable element that could fall under existing intellectual property protections, though significant gaps remain. The Resemble AI report underscores that while some jurisdictions are moving toward explicit voice rights, others rely on piecemeal applications of privacy, trademark, or personality laws. The convergence of these efforts suggests an emerging global consensus: voice cloning without consent constitutes a violation worthy of legal remedy, but the mechanisms remain inconsistent and evolving.

Mexico's Platform Liability for Deepfakes

Global lawmakers are scrambling to catch up with AI voice cloning technology that can replicate a person’s speech with unsettling accuracy. In Mexico, legislators have taken the boldest step yet by voting to impose liability on digital platforms that host unauthorized AI-generated voices, forcing them to remove such content promptly or face legal consequences. This shift treats voice as a protected property right, akin to image or trademark, and places the burden of enforcement on intermediaries rather than individuals who must first prove harm. Other jurisdictions are watching closely, recognizing that platform-centric rules may be more practical than trying to regulate every creator directly.

In Europe, Lithuania has proposed criminalizing the use of cloned voices without consent, while the United Kingdom explores whether existing privacy and fraud statutes can cover synthetic speech. Australia is examining copyright law as a possible shield, arguing that a voice could be considered a protectable performance. Across these efforts, a common tension emerges: how to safeguard personal identity without stifling innovation or free expression. The debate is no longer theoretical—voices are already being cloned for impersonation, disinformation, and entertainment—and the law is racing to decide who owns a sound that no human ever truly made.

Industry Self-Regulation vs Government Oversight

Global lawmakers are scrambling to catch up with the rapid spread of AI voice cloning technology that can replicate a person’s voice with unsettling accuracy. Lithuania has emerged as one of the first European nations to propose dedicated legal safeguards, with lawmakers drafting legislation that would explicitly criminalize unauthorized cloning of a citizen’s voice and require platforms to remove such content upon request. Across the Atlantic, Mexico’s congress has taken a different approach by voting to impose takedown obligations on digital platforms, compelling them to pull down audio files that feature AI-altered voices without consent, a move that mirrors broader content moderation frameworks already applied to images and text.

In the United Kingdom, however, the legal landscape remains fragmented, with experts warning that existing copyright and data protection laws may offer insufficient protection against sophisticated voice synthesis. Australia’s legal scholars are exploring whether voice itself can be treated as a form of intellectual property under copyright law, while industry players like Resemble AI are publishing transparency reports and watermarking techniques in a bid to self-regulate. The tension between voluntary industry standards and binding government mandates is intensifying as courts begin to confront the first test cases involving impersonation, fraud, and reputational harm caused by cloned voices.

Global AI Voice Cloning Laws at a Glance

Global legal responses to AI voice cloning remain fragmented, with most nations relying on existing privacy, copyright, or personality rights laws. Lithuania and Mexico are pioneering targeted legislation, while the UK and Australia explore reforms. Without consent, voice cloning risks eroding individual identity and trust, urging urgent, harmonized legal frameworks.