Stolen Voices: The Hidden Crisis of AI Voice Cloning Without Consent

Abstract

A voice is the most intimate signature a human being carries more personal than a signature on paper, more immediate than a photograph. It carries a mother’s worry, a grandfather’s warmth, a celebrity’s charisma, a leader’s authority. For most of human history, a voice could not be separated from the body that produced it. That certainty has now collapsed. With as little as three seconds of audio, modern AI voice-cloning systems can reproduce a person’s pitch, cadence, breath pattern, and emotional inflection with unsettling accuracy. What was once the stuff of science fiction, a machine that can make anyone say anything is now a browser tab away from anyone with a credit card and bad intentions.

This article examines the phenomenon of non-consensual AI voice cloning: how it works, who it harms, and why the law is scrambling to catch up. It draws on real courtroom battles from Amitabh Bachchan and Karan Johar in India to Scarlett Johansson’s clash with OpenAI and a fabricated Joe Biden robocall that shook a U.S. presidential primary alongside harrowing personal stories of families terrorized by fake kidnapping calls. It analyzes the legal doctrines being stretched to their limits (personality rights, publicity rights, privacy law, and emerging AI statutes), evaluates current policy responses across India, the United States, and the European Union, and proposes a framework of technical, legal, and social safeguards. It closes not with despair, but with a case for vigilance because in an age when anyone’s voice can be stolen, protecting the human voice is no longer a luxury. It is a necessity.

 

Introduction: The Sound of Betrayal

Imagine picking up the phone and hearing your daughter’s voice, shaking, sobbing, saying, “Mom, I messed up.” You would not stop to ask whether the voice was real. No parent does. That is exactly what happened to Jennifer DeStefano, an Arizona mother, in early 2023. The voice on the line belonged, in every detail – inflection, pitch, even the specific way her daughter cried. Except her daughter was safely skiing, a hundred miles away, and had never made the call at all. A criminal had cloned her voice using a few seconds of audio scraped from somewhere online, then used it to extort a terrified mother out of tens of thousands of dollars. DeStefano bought time, verified her daughter was safe within four minutes, and escaped financial loss. Millions of others facing similar calls are not so fortunate.

This is the quiet, chilling underside of one of the most celebrated breakthroughs in artificial intelligence. Voice-cloning and text-to-speech technology have genuinely wonderful uses: restoring speech to people who have lost their voices to illness, dubbing films into dozens of languages without losing an actor’s original performance, letting a grandmother who has passed away still “read” a bedtime story to her grandchildren from an old recording. But every technology that can comfort can also be turned into a weapon, and voice cloning is no exception. The same tool that can give a stroke survivor their voice back can also let a scammer impersonate a company’s CFO and authorize a fraudulent wire transfer, or let a stranger generate a fake political speech designed to manipulate an election, or let an anonymous troll put obscene words into a public figure’s mouth for viral clicks.

What makes this moment different from earlier waves of digital fraud is speed and scale. Phishing emails require some effort to fall for. A cloned voice bypasses our rational filters entirely, it exploits a cognitive shortcut millions of years in the making: if it sounds like someone we trust, we believe it is them. As Arizona State University AI researcher Subbarao Kambhampati put it after the DeStefano case, “You can no longer trust your ears.” That single sentence captures the crisis at the heart of this article.

Over the following sections, we will walk through the technology, the human cost, the courtroom battles being fought right now on multiple continents, the gaps in the law, and, most importantly, what can actually be done about it.

 

Part 1: How AI Voice Cloning Actually Works

Modern voice cloning relies on deep generative models, neural networks trained on vast quantities of recorded speech that learn the relationship between text and the acoustic patterns of a specific voice. Early systems required hours of clean audio to produce a passable clone. Today’s commercial tools need only a few seconds. A voicemail greeting, a wedding toast uploaded to social media, a YouTube video, a viral Instagram reel, a customer service call, any of these can become raw material.

The process generally works in three stages:

1. Extraction – audio samples of the target voice are collected, often without their knowledge, from social media, podcasts, interviews, or even a stolen phone call.
2. Training/Embedding – the AI model converts that audio into a compact “voiceprint,” a mathematical representation of the person’s unique vocal characteristics.
3. Synthesis – new text is fed into the model, and it generates speech in the cloned voice, complete with emotional tone, accent, and pacing that can be adjusted to sound panicked, angry, cheerful, or persuasive.

The frightening part is not the technology’s sophistication, it is its accessibility. What once demanded a research lab and specialized expertise is now available through consumer apps, some free, some for a few dollars a month. This democratization is precisely why the harms discussed in this article have moved so quickly from theoretical to widespread.

Part 2: Stories That Should Keep Us Up at Night

The Mother Who Heard Her Daughter Scream

Jennifer DeStefano’s ordeal, described above, is now one of the most cited voice-cloning cases in America precisely because it illustrates how ordinary families, not just celebrities, are targets. The FBI has noted that such “virtual kidnapping” calls, once reliant on generic threats, are increasingly built around cloned voices of real children, and that families in the United States lose an average of roughly $11,000 per incident when scams like this succeed. The trauma inflicted in those four minutes, however, cannot be measured in dollars. DeStefano has said the experience permanently changed how she answers her phone, she is now wary of unknown numbers and has considered that a scammer may have captured her daughter’s voice from something as mundane as an unanswered “hello” on a prank call.

The Grandparent Scam, Reimagined

Long before AI, criminals ran the classic “grandparent scam”, a caller pretending to be a grandchild in trouble, asking for emergency bail or hospital money. Voice cloning has supercharged this old con. Instead of a stranger’s voice roughly approximating a young adult in distress, elderly victims now hear what is, to their ears, unmistakably their own grandchild. Consumer protection agencies and elder-law attorneys increasingly warn that this is one of the fastest-growing forms of elder financial abuse in the AI era, precisely because it weaponizes love rather than greed.

When a Presidential Primary Got a Robocall From a Dead Man’s Understudy, Almost

In January 2024, thousands of voters in New Hampshire picked up their phones to hear what sounded exactly like President Joe Biden, urging them to “save your vote” and skip the state’s presidential primary. It was entirely fake, a cloned voice generated by political consultant Steve Kramer, who later admitted to commissioning it. The Federal Communications Commission ultimately imposed a $6 million fine against Kramer for the illegal robocalls, and the telecom carrier that transmitted the spoofed calls, Lingo Telecom, separately paid a $1 million settlement for failing basic caller-verification protocols. Kramer was also indicted on state charges in New Hampshire, though a jury later acquitted him of the criminal counts in 2025. Whatever the ultimate legal outcome for Kramer personally, the episode did what no white paper could: it proved, in the middle of a live democratic process, that a fabricated voice could reach thousands of citizens and attempt to manipulate how they exercised their most fundamental civic right.

The Actress Who Said No, and Heard Herself Anyway

When OpenAI unveiled a conversational voice for ChatGPT named “Sky” in 2024, listeners immediately noted its eerie similarity to actress Scarlett Johansson, an irony not lost on anyone familiar with her role voicing an AI companion in the film Her. Johansson has said that Sam Altman had approached her about voicing the assistant and that she declined; when “Sky” launched sounding remarkably like her anyway, she raised concerns about consent and the unauthorized use of her likeness. OpenAI paused the voice and stated the resemblance was unintentional, but the episode became a flashpoint in the global conversation about whether “sounding like” someone without using their actual recorded voice can itself be a form of appropriation, and it renewed calls from actors, especially those represented by unions such as SAG-AFTRA, for enforceable consent requirements around AI-generated voice performances.

Bollywood’s Reckoning: When Superstars Went to Court to Reclaim Their Own Voices

Nowhere has the courtroom battle over voice cloning been more vivid, or more instructive, than in India, where a wave of Bollywood stars turned to the Delhi and Bombay High Courts starting in 2022.

• Amitabh Bachchan v. Rajat Nagi (2022): The legendary actor discovered that his name, image, voice, and even his famous catchphrases were being exploited across websites, prank-call apps, merchandise, and domain names without his consent, some of it framed as “Amitabh Bachchan video calls.” The Delhi High Court granted what became one of India’s first detailed “John Doe” orders in this space, an injunction issued against unnamed, unknown defendants “at large,” designed to automatically cover future infringers, including uses in emerging mediums like NFTs and the metaverse. It was an early, landmark recognition that a celebrity’s voice is not free raw material for anyone with a microphone and bad intentions.
• Anil Kapoor v. Simply Life India (2023): The Delhi High Court restrained more than a dozen entities from using actor Anil Kapoor’s name, image, and voice, including his signature catchphrase “Jhakaas”, through AI, machine learning, deepfakes, and face-morphing tools, whether for profit or not. The court drew a careful line: parody and satire remain protected free speech, but content designed to tarnish or commercially exploit a person’s identity is not.
• Arijit Singh v. Codible Ventures LLP (2024): In what is widely regarded as India’s first judgment squarely addressing generative-AI voice cloning, playback singer Arijit Singh, one of the most-streamed artists in the world, sued a company that was using AI to synthesize his voice and falsely suggest his endorsement of a virtual event. The Bombay High Court ruled that his voice, vocal style, and mannerisms were protected personality attributes and issued an injunction covering not just current platforms but future technologies, including voice-conversion tools, digital avatars, and the metaverse, a forward-looking, “dynamic” injunction designed to stay effective as the underlying technology evolves.
• Karan Johar v. Ashok Kumar & Ors. (2025): Filmmaker Karan Johar approached the Delhi High Court after AI tools were used to create morphed, sexually suggestive, and abusive deepfake content bearing his name, image, voice, and his famous nickname “KJo.” The court granted an ex-parte injunction restraining not only the direct infringers but also directing intermediary platforms, including domain registrars, e-commerce sites, and GIF-hosting platforms, to take down infringing content and preserve user data for enforcement, while explicitly preserving room for genuine satire and fair comment.
• A Cascade of Cases in 2025: The Johar ruling was part of a broader surge. Aishwarya Rai Bachchan sued over a website exploiting her name and image; Abhishek Bachchan brought action against merchandise sellers and AI platforms depicting him in derogatory contexts; even Aaradhya Bachchan, a minor, was the subject of a case after false, sensationalized videos about her health circulated on YouTube. Veteran singers Asha Bhosle and Kumar Sanu, sitarist Ravi Shankar’s estate, and actor Suniel Shetty followed with their own suits against AI-driven impersonation.

Together, these cases represent the fastest-evolving body of “personality rights” jurisprudence anywhere in the world, built not through a single statute, but case by case, judge by judge, because India, like most countries, still has no dedicated law for AI-era identity theft.

The Voice That Was Never Consented To, and Never Will Age

There is a quieter, more philosophical case worth noting too: the question of the dead. Legal experts examining India’s personality-rights cases have pointed out that publicity and privacy rights are not currently inheritable there, meaning a deceased singer’s estate may have no clear legal recourse if an AI recreates their voice for a new “posthumous” song or advertisement. The same debate is unfolding globally, from AI-generated “new” tracks in the voice of deceased musicians to synthetic recreations of actors who can no longer consent to a performance. It is a reminder that voice cloning does not just threaten the living, it threatens the right of the dead to rest.

 

Part 3: Legal Analysis, A Patchwork Straining to Hold

India: Judge-Made Law Filling a Legislative Vacuum

India has no standalone statute for “personality rights” or “voice rights.” Instead, courts have stitched together protection from constitutional privacy jurisprudence (notably the landmark K.S. Puttaswamy privacy judgment), common-law torts like passing off, trademark principles, the Information Technology Act’s provisions on intermediary takedown, and the general “right of publicity” first recognized in cases like ICC Development v. Arvee Enterprises (2003) and Titan Industries v. Ramkumar Jewellers (2012). To win, a celebrity plaintiff typically must show three things: (1) they have identifiable celebrity status, (2) the public can recognize them from the unauthorized use, and (3) the use serves someone else’s commercial gain. The 2022–2025 wave of AI-specific cases has extended this framework to cover deepfakes, voice conversion, GIFs, AI chatbots, and even the metaverse, but it remains, fundamentally, an ad hoc judicial patch rather than a codified right, and India’s proposed Digital India Act, still under consultation, is expected to be the first real attempt at closing that gap.

The United States: A Fifty-State Quilt

The U.S. has no federal right of publicity; instead, each state defines its own version, and many say nothing about AI-generated voices at all. Some states have moved fast, Tennessee’s ELVIS Act (2024), named in tribute to Elvis Presley, was the first U.S. law to explicitly protect a person’s voice from unauthorized AI simulation, closing a gap that older right-of-publicity statutes never anticipated. At the federal level, momentum has built behind the bipartisan NO FAKES Act, which would create a nationwide right against unauthorized AI-generated replicas of a person’s voice or likeness, though as of this writing it has not been enacted. Meanwhile, the FCC has used its existing authority over robocalls to declare AI-generated voices in unsolicited calls illegal under the Telephone Consumer Protection Act, the legal hook that made the Kramer prosecution possible, while the FTC pursues voice-cloning scams under its general consumer-protection mandate.

The European Union: Regulating the Machine, Not Just the Harm

The EU’s approach is different in character: rather than building a new “voice right,” the EU AI Act classifies certain deepfake and biometric-categorization uses as high-risk or outright prohibited, and imposes transparency obligations requiring that AI-generated or manipulated audio, image, and video content be clearly labeled as such. This sits alongside the GDPR, which treats a person’s voice as biometric or personal data in many contexts, giving individuals a data-protection hook against unauthorized processing, a fundamentally different legal theory than the publicity-rights model used in India and parts of the U.S.

The Common Thread, and the Common Gap

Every jurisdiction examined here shares one structural weakness: the law was built for a world where impersonation required a human being who could be identified, located, and sued. AI voice cloning breaks that assumption. The “John Doe” orders pioneered in the Bachchan and Johar cases are a direct judicial response to this, they let courts act against anonymous, unknown, even not-yet-identified infringers, and against platforms and intermediaries that host the harm, because chasing down each individual bad actor is often impossible in practice. Expect more courts worldwide to lean on this model, and expect legislators to eventually have to follow where judges have already gone.

 

Part 4: The Deeper Harm, Beyond Money and Merchandise

It would be a mistake to file this issue away as a subcategory of financial fraud or celebrity branding disputes. The Delhi High Court itself, in the Karan Johar matter, explicitly linked non-consensual voice and image manipulation to a violation of dignity and privacy under Article 21 of the Indian Constitution, the same provision that guarantees the right to life. That framing matters. A cloned voice can:

• Terrorize families, as in virtual kidnapping scams, causing acute psychological trauma even when no money is lost.
• Defraud the vulnerable, particularly the elderly, who are disproportionately targeted by “grandparent scams” and are often less familiar with the existence of the technology being used against them.
• Corrode democratic trust, as the Biden robocall demonstrated, a single fabricated call reached thousands of voters with a message designed to suppress turnout.
• Enable sexual and reputational abuse, as seen when deepfake audio and video are combined to produce obscene or humiliating content bearing a real person’s voice, something both Karan Johar and Aishwarya Rai Bachchan specifically had to fight in court.
• Undermine artistic livelihood, particularly for voice actors, playback singers, and dubbing artists whose entire professional identity, their voice, can now be replicated without payment or credit, a concern central to Arijit Singh’s case and to SAG-AFTRA’s advocacy in the United States.
• Erode the very reliability of human testimony. Courts, journalists, and ordinary citizens have long treated audio recordings as strong evidence. Once convincing fake audio is trivial to produce, every real recording becomes deniable, a phenomenon researchers call the “liar’s dividend,” where wrongdoers can dismiss genuine evidence against them as “just an AI clone.”

This last point deserves emphasis: the danger of voice cloning is not only that fake things will be believed, but that real things will be disbelieved. That is a corrosive, society-wide cost that no single lawsuit can repair.

 

Part 5: Recommendations and Policy Directions

No single fix will solve this. What follows is a layered framework, legal, technical, corporate, and personal, because a threat this distributed needs a defense this distributed.

1. Legal and Regulatory Reform

• Enact dedicated, technology-neutral voice and likeness rights. Laws like Tennessee’s ELVIS Act and the proposed federal NO FAKES Act in the U.S. should become templates globally, explicitly covering AI-generated and AI-cloned voices, not just traditional impersonation.
• Extend protection beyond celebrities. Most existing personality-rights litigation involves the famous, because they have the resources to sue. Ordinary citizens, the DeStefanos of the world, need accessible, low-cost legal remedies too, not just criminal fraud statutes that only apply after money has already been lost.
• Mandate AI-content labeling and provenance standards, following the EU AI Act’s transparency requirements, so that synthetic audio can be technically flagged and traced back to its generating platform.
• Close the posthumous gap. Countries should clarify whether voice and likeness rights survive death and, if so, for how long and who inherits them, an issue currently unresolved in India and many other jurisdictions.
• Hold platforms and telecom carriers accountable, as the FCC did with Lingo Telecom, for failing to implement basic verification safeguards that could catch spoofed and AI-generated calls before they reach victims.

2. Technical Safeguards

• Watermarking and provenance tracking embedded at the point of generation, so AI-made audio carries a detectable signature even after editing or compression.
• Consent-verification layers built into commercial voice-cloning platforms, requiring proof that the voice being cloned belongs to the user or was licensed, an approach some responsible AI voice companies have already begun adopting after early controversies over misuse of their tools.
• Detection tools for consumers and institutions, banks, courts, and newsrooms in particular, to verify whether audio is synthetic before acting on it.

3. Corporate and Institutional Responsibility

• AI companies should build default restrictions against cloning any voice without explicit, verifiable consent from the person being cloned, not opt-out policies enforced only after public backlash, as happened with OpenAI’s “Sky” voice.
• Entertainment guilds and unions should keep negotiating enforceable consent and compensation clauses for any AI use of a performer’s voice, extending the protections U.S. actors and voice artists won in recent industry negotiations.
• Financial institutions and call centers should adopt mandatory callback verification protocols for any high-value transaction requested by phone, regardless of how certain the voice on the line sounds.

4. Personal and Family-Level Protection

• Establish a family “safe word”, a phrase never posted online, that can instantly verify identity during a suspicious emergency call, exactly as the FBI recommended after the DeStefano case.
• Limit publicly posted audio and video of children and vulnerable family members where feasible, since even a few seconds of clear audio is enough to build a clone.
• Pause before paying. If a call demands urgent payment through hard-to-trace methods, wire transfers, cryptocurrency, gift cards, treat it as an immediate red flag, regardless of how convincing the voice sounds, and independently call the person back on their known number.
• Normalize doubt as a form of care, not disloyalty. Verifying that a loved one is truly safe is not an insult to them; it is the single most effective defense that currently exists against this crime.

5. Public Awareness and Digital Literacy

Perhaps the most underrated intervention is simply widespread public knowledge that this technology exists and is cheap. Every documented case in this article, from Bollywood courtrooms to a dance studio in Scottsdale, shows that people who did not know voice cloning was possible were the most vulnerable to it. Governments, schools, and media literacy programs should treat “you can no longer trust your ears” as a lesson as fundamental as “don’t click suspicious links” once was.

 

Conclusion

There is something almost unbearably tender about a human voice. It is how a mother is recognized by her child before the child can even see clearly. It is how we say “I love you,” “I’m sorry,” and “I’m scared” in ways no text message can replicate. Jennifer DeStefano did not doubt the voice on the phone for a single second, not because she was careless, but because a mother knows her child’s voice the way she knows her own heartbeat. That instinct, built over millennia of evolution and years of listening to a child grow, was hijacked in seconds by a stranger with a laptop.

And yet, in courtroom after courtroom, from the Delhi High Court defending Amitabh Bachchan’s decades of hard-earned reputation, to Arijit Singh insisting that the voice that moved millions of listeners still belongs to him, to Karan Johar refusing to let his name and nickname be twisted into cruelty, to a jury in New Hampshire wrestling with what democracy owes its voters when even a president’s voice can be faked, there is also something quietly hopeful. People are fighting back. Judges are inventing new remedies faster than legislatures can pass old ones. Families are learning new instincts, like safe words, that their grandparents never needed. Technologists are beginning, however belatedly, to build consent into the very tools that once ignored it.

The voice is the last truly personal thing many of us have left in a world where our faces, our data, and our words are already scattered across the internet. It deserves the same fierce protection we would give a fingerprint, a signature, a heartbeat. Because the next voice on the other end of a stranger’s phone call may not be your daughter, your grandfather, or your favorite singer at all. It may simply be a machine, wearing the sound of someone you love like a mask. The question this technology forces on all of us is not whether we can build it. We already have. The question is whether we will build, just as urgently, the laws, the tools, and the habits of doubt that let us keep believing in the voices that are real.

 

Frequently Asked Questions (FAQ)

Q1. What exactly is AI voice cloning? It is the use of machine-learning models trained on samples of a person’s speech to generate new, artificial audio that mimics their voice, tone, and speech patterns, sometimes with just a few seconds of source audio.

Q2. Is AI voice cloning illegal? It depends on jurisdiction, consent, and intent. Using someone’s cloned voice without consent for fraud, defamation, harassment, or unauthorized commercial gain is illegal or actionable in most countries through fraud law, personality/publicity rights, privacy law, or specific statutes like Tennessee’s ELVIS Act. Voice cloning for consented, licensed, or clearly disclosed purposes (dubbing, accessibility tools, parody protected as free speech) is generally lawful.

Q3. How much audio does someone need to clone a voice? Some commercial tools now require as little as three to ten seconds of clear audio, which can be pulled from social media videos, voicemail greetings, interviews, or even a brief phone call.

Q4. How can I tell if a call or recording uses a cloned voice? It is increasingly difficult by ear alone. Warning signs include unusual urgency, requests for untraceable payment methods, resistance to being called back, slightly flat emotional transitions, or background sound that doesn’t match the claimed situation. The most reliable defense is independent verification, hanging up and calling the person back on a known number.

Q5. What should I do if I receive a suspicious call using a cloned voice? Stay calm, avoid confirming personal information, try to contact the person independently through another channel, and report the incident to local law enforcement and, in the U.S., the FTC or FCC. Do not send money or gift cards under pressure.

Q6. Can celebrities really stop people from cloning their voices? Increasingly, yes, at least through injunctions. Courts in India (in cases involving Amitabh Bachchan, Arijit Singh, Anil Kapoor, and Karan Johar, among others) have granted broad injunctions, including against unknown future infringers, restraining unauthorized AI use of a celebrity’s voice, image, and likeness. Enforcement across borders and against anonymous actors, however, remains genuinely difficult.

Q7. Do ordinary, non-famous people have any legal protection? Generally less than celebrities, since most existing case law centers on commercial exploitation of famous personas. However, fraud statutes, harassment law, data protection law (like the GDPR’s treatment of voice as biometric data), and newer AI-specific laws increasingly offer protection to private individuals too, particularly in scam and harassment cases.

Q8. What is a “John Doe” order, and why does it matter here? It is a legal order issued against unknown, unnamed defendants, “the world at large”, rather than a specific identified person. Indian courts have used this tool extensively in voice-cloning and deepfake cases because the anonymous, distributed nature of online infringement often makes it impossible to identify every wrongdoer in advance.

Q9. Are there legitimate, ethical uses of voice cloning? Yes, assistive technology for people who have lost their voice to illness (such as ALS or throat cancer), multilingual film dubbing with the original actor’s consent, audiobook narration, and accessibility tools all rely on licensed, consented voice synthesis. The ethical line is consent, not the technology itself.

Q10. What is being done at the policy level to address this? Responses vary by region: India is developing case law through the courts and considering a broader Digital India Act; the U.S. has state laws like the ELVIS Act and a proposed federal NO FAKES Act, alongside FCC and FTC enforcement against AI robocalls and scams; and the EU regulates through the AI Act’s transparency and risk-based rules alongside GDPR’s biometric data protections.

 

REFERENCES, BLUEBOOK (21st ED.) FORMAT

1. Constitutional and Statutory Materials
2. The Constitution of India art. 21.
3. Information Technology Act, No. 21 of 2000, India Code (2000).
4. Digital Personal Data Protection Act, No. 22 of 2023, India Code (2023).
5. Trade Marks Act, No. 47 of 1999, India Code (1999).
6. Telephone Consumer Protection Act of 1991, 47 U.S.C. § 227 (2018).
7. Code Ann. § 47-25-1101 to -1108 (Ensuring Likeness, Voice, and Image Security (“ELVIS”) Act, 2024),
8. Nurture Originals, Fostering Art, and Keeping Entertainment Safe Act (“NO FAKES Act”), S. 4875, 118th Cong. (2024)
9. Regulation 2024/1689, of the European Parliament and of the Council of 13 June 2024 Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act), 2024 O.J. (L 1689).
10. Regulation 2016/679, of the European Parliament and of the Council of 27 April 2016 on the Protection of Natural Persons with Regard to the Processing of Personal Data (General Data Protection Regulation), 2016 O.J. (L 119) 1.
11. Case Law (India)
12. Justice K.S. Puttaswamy (Retd.) v. Union of India, (2017) 10 SCC 1.
13. ICC Development (International) Ltd. v. Arvee Enterprises, 2003 (26) PTC 245 (Del.).
14. Titan Indus. Ltd. v. Ramkumar Jewellers, 2012 SCC OnLine Del 2382.
15. Amitabh Bachchan v. Rajat Nagi, CS(COMM) 819/2022, 2022 SCC OnLine Del 4110 (Del. H.C.).
16. Anil Kapoor v. Simply Life India, CS(COMM) 652/2023, 2023 SCC OnLine Del 6914 (Del. H.C.).
17. Arijit Singh v. Codible Ventures LLP, 2024 SCC OnLine Bom 2445 (Bom. H.C. July 26, 2024).
18. Karan Johar v. India Pride Advisory Pvt. Ltd., COMM IPR Suit (L) No. 17863 of 2024 (Bom. H.C.).
19. Karan Johar v. Ashok Kumar, CS(COMM) [no.], 2025 SCC OnLine Del 6108 (Del. H.C. Sept. 17, 2025).
20. Aaradhya Bachchan v. YouTube Channels, CS(COMM) [no.] (Del. H.C. 2023) (exact suit number/reporter citation unverified).
21. Aishwarya Rai Bachchan v. Aishwaryaworld.com, CS(COMM) 956/2025, 2025 SCC OnLine Del 5943 (Del. H.C.).
22. Abhishek Bachchan v. Bollywood Tee Shop, CS(COMM) [no.] (Del. H.C. 2025) (exact suit number/reporter citation unverified).
23. Ravi Shankar v. John Doe(s), 2025 SCC OnLine Del 6332 (Del. H.C.).
24. Suniel V. Shetty v. John Doe(s), I.A. No. 32198 of 2025 (Bom. H.C.).

III. Regulatory and Enforcement Materials (United States)

1. Commc’ns Comm’n, Notice of Apparent Liability for Forfeiture, In re Steven
2. Kramer, File No. EB-TCD-24-00036850 (May 23, 2024).
3. Commc’ns Comm’n, Forfeiture Order, In re Steven Kramer, File No. EB-TCD-24-00036850 (Sept. 26, 2024).
4. Commc’ns Comm’n, Consent Decree, In re Lingo Telecom, LLC, File No. EB-TCD-24-00036851 (Aug. 21, 2024).
5. Trade Comm’n, Consumer Alerts on AI Voice Cloning Scams (2023–2024) (available at ftc.gov).
6. Government Materials, Advisories, and Reports
7. Bureau of Investigation, Public Service Announcement, Criminals Use Generative
8. Artificial Intelligence to Facilitate Financial Fraud (June 2024).
9. H. Att’y Gen. Office, Press Release on Indictment of Steven Kramer (May 23, 2024).
10. Secondary Sources
11. WIPO Magazine, AI Voice Cloning: How a Bollywood Veteran Set a Legal Precedent (Apr. 17, 2025), https://www.wipo.int/en/web/wipo-magazine/articles/ai-voice-cloning-how-a-bollywood-veteran-set-a-legal-precedent-73631.
12. SCC Online Blog, Delhi HC Protects Karan Johar’s Personality Rights (Sept. 22, 2025), https://www.scconline.com/blog/post/2025/09/22/delhi-hc-protects-karan-johars-personality-rights/.
13. Chambers & Partners, Personality Rights and the Bachchan Cases: A New Chapter in Indian Jurisprudence (Nov. 6, 2025), https://chambers.com/articles/personality-rights-and-the-bachchan-cases-a-new-chapter-in-indian-jurisprudence.
14. S. Rana & Co., Personality Rights and the Bachchan Cases: A New Chapter in Indian Jurisprudence (Dec. 23, 2025), https://ssrana.in/articles/personality-rights-and-the-bachchan-cases-a-new-chapter-in-indian-jurisprudence/.
15. Cyril Amarchand Mangaldas, Personality Rights in India in the Age of AI, Disp. Resol. Blog (May 14, 2026), https://disputeresolution.cyrilamarchandblogs.com/2026/05/personality-rights-in-india-in-the-age-of-ai/.
16. IPRMENTLAW, AI Voice Cloning and Personality Rights: A New Challenge for the Music Industry (Jan. 14, 2024), https://iprmentlaw.com/2024/01/14/ai-voice-cloning-and-personality-rights-a-new-challenge-for-the-music-industry/.
17. IIPRD, The Mirage of Fame: Deepfakes, AI and Evolving Jurisprudence on Personality Rights (Dec. 2, 2025), https://www.iiprd.com/the-mirage-of-fame-deepfakes-ai-and-evolving-jurisprudence-on-personality-rights/.
18. World Trademark Review (Lexology), Bollywood Singer Prevails in First AI Voice-Cloning Infringement Decision in India (Aug. 29, 2024), https://www.lexology.com/library/detail.aspx?g=c3a5b346-9c4e-40db-9672-da89dc0fbba4.
19. Ruhi Khan, From Deepfakes to Dignity: What Bollywood’s Personality Rights Battle with AI Tells Us, LSE Media@LSE Blog (Oct. 14, 2025), https://blogs.lse.ac.uk/medialse/2025/10/14/from-deepfakes-to-dignity-what-bollywoods-personality-rights-battle-with-ai-tells-us/.
20. Record of Law, Deepfakes and the Law: Addressing Legal Accountability for AI-Generated Misinformation (Dec. 7, 2025), https://recordoflaw.in/deepfakes-and-the-law-addressing-legal-accountability-for-ai-generated-misinformation/.
21. The IP Press, Lights, Camera, Lawsuits: 2025’s Personality Rights Explosion (Jan. 30, 2026), https://www.theippress.com/2026/01/30/lights-camera-lawsuits-2025s-personality-rights-explosion/.
22. Commc’ns Comm’n, FCC Issues $6M Fine for N.H. Robocalls (Sept. 26, 2024), https://www.fcc.gov/document/fcc-issues-6m-fine-nh-robocalls.
23. Bobby Allyn, Criminal Charges and FCC Fines Issued for Deepfake Biden Robocalls, NPR (May 23, 2024), https://www.npr.org/2024/05/23/nx-s1-4977582/fcc-ai-deepfake-robocall-biden-new-hampshire-political-operative.
24. Derek B. Johnson, FCC Hits Operative Behind New Hampshire Robocall with $6 Million Fine, CyberScoop (Sept. 26, 2024), https://cyberscoop.com/fcc-fine-joe-biden-deepfake-new-hampshire-robocall-steve-kramer/.
25. H. Pub. Radio, Political Operative Who Admitted to Creating Fake Biden Robocalls Found Not Guilty (June 16, 2025), https://www.nhpr.org/nh-news/2025-06-13/political-operative-fake-biden-robocalls-nh-primary-found-not-guilty.
26. Julia Jacobo, AI Scam Calls: This Mom Believes Fake Kidnappers Cloned Her Daughter’s Voice, CNN (Apr. 29, 2023), https://www.cnn.com/2023/04/29/us/ai-scam-calls-kidnapping-cec/index.html.
27. AZFamily/KPHO, ‘I’ve Got Your Daughter’: Scottsdale Mom Warns of Close Call with AI Voice Cloning Scam (Apr. 10, 2023), https://www.azfamily.com/2023/04/10/ive-got-your-daughter-scottsdale-mom-warns-close-encounter-with-ai-voice-cloning-scam/.
28. Global News, AI Kidnapping Scam Copied Teen Girl’s Voice in $1M Extortion Attempt (Apr. 18, 2023), https://globalnews.ca/news/9629883/ai-kidnapping-scam-teen-girl-voice-cloned-extortion-arizona-jennifer-destefano/.
29. Bar Ass’n, A True Story of a Grandparent “Kidnapping” Scam, Voice of Experience (Mar. 14, 2026), https://www.americanbar.org/groups/senior_lawyers/resources/voice-of-experience/2026-march/grandparent-kidnapping-scam/.
30. Andrew R. Chow, AI Voice Cloning Is on the Rise. Here’s What to Know, TIME (Apr. 13, 2026), https://time.com/6275794/ai-voice-cloning-scams-music/.
31. OpenTools AI News, Scarlett Johansson Takes on OpenAI: The Voice Cloning Controversy Unfolds (Mar. 16, 2025), https://opentools.ai/news/scarlett-johansson-takes-on-openai-the-voice-cloning-controversy-unfolds.