Blab AI
Delhi Police Arrest Five Men in 10-Crore AI Voice-Cloning Scam Targeting Mumbai Company
On a rainy Tuesday in Delhi, law‑enforcement officials shut the curtain on a high‑stakes fraud that crossed state lines and exploited the very human element of trust. Five men—Vikas, Vansh, Faiyaz Alam, Amit and Balveer Kumar—were taken into custody on 17 June 2026 after police said they used artificial‑intelligence (AI) voice‑cloning technology to impersonate the managing director (MD) of a Mumbai‑based firm and swindle its deputy general manager (DGM) out of more than ₹10 crore.
The accused created a synthetic voice that matched the MD’s speech patterns with uncanny precision. They called the DGM and, under the pretense of an urgent order from the MD, persuaded him to transfer funds without the usual verification checks. Investigators say the attackers leveraged the company’s internal trust hierarchy and the urgency of the request to override standard safeguards.
Once the money left the company’s accounts, the perpetrators layered it through 63 different bank accounts—a common technique in financial crimes that obfuscates the origin of the funds and delays detection. The total amount transferred exceeded ₹10 crore.
The operation began to unravel when Vikas was detained at a bank while attempting to withdraw cheque‑linked funds. Bank staff reported suspicious activity, prompting police intervention. During interrogation, Vikas supplied details that led Delhi Police to identify and arrest the other four suspects. Police say each accused had a specific role, from technical support for the voice‑cloning software to managing the financial transfers.
Mumbai Police had already registered a cyber‑fraud case after the company lodged a formal complaint. As the scale of the scheme became clearer, Delhi Police joined the investigation. The two agencies are now collaborating to trace the full financial trail, identify additional beneficiaries, and determine whether the suspects belong to a larger cyber‑crime network operating across state lines.
Voice‑cloning, also known as deepfake audio, uses AI models to generate speech that mimics a target individual. The technology can produce realistic voice outputs from only a few seconds of audio. While originally developed for entertainment and accessibility applications, it has become increasingly accessible to the public and can be misused for social‑engineering attacks.
Cybersecurity experts note that such scams rely on manipulating human decision‑making rather than hacking systems directly. By convincing employees that a senior executive is issuing orders, attackers bypass technical controls. Experts advise organisations to adopt stronger safeguards, such as multi‑factor authentication for financial approvals, call‑back verification procedures, and employee training on deepfake awareness.
The case underscores the growing threat of AI‑driven financial fraud in India. According to reports, the use of synthetic media for scams has risen sharply in 2026, with several incidents involving voice‑cloning and deepfake video. The Delhi‑Mumbai scam is one of the first high‑profile cases where AI was used to orchestrate a large‑scale corporate fraud.
At present, the five suspects remain in custody, and Delhi and Mumbai Police are continuing to investigate the possibility of a wider network. The company that was defrauded has not released a statement, but industry observers expect it to review its internal controls and consider implementing voice‑authentication technologies.
The incident highlights the need for coordinated law‑enforcement efforts to tackle technologically sophisticated crimes that cross state borders. It also raises questions about regulatory frameworks for AI tools that can be weaponised for fraud.
As of 19 June 2026, the investigation is ongoing. No further arrests have been announced, and the company has not confirmed the recovery of the stolen funds. The case remains a cautionary example of how AI can be exploited to undermine trust and financial integrity in corporate settings.
The accused created a synthetic voice that matched the MD’s speech patterns with uncanny precision. They called the DGM and, under the pretense of an urgent order from the MD, persuaded him to transfer funds without the usual verification checks. Investigators say the attackers leveraged the company’s internal trust hierarchy and the urgency of the request to override standard safeguards.
Once the money left the company’s accounts, the perpetrators layered it through 63 different bank accounts—a common technique in financial crimes that obfuscates the origin of the funds and delays detection. The total amount transferred exceeded ₹10 crore.
The operation began to unravel when Vikas was detained at a bank while attempting to withdraw cheque‑linked funds. Bank staff reported suspicious activity, prompting police intervention. During interrogation, Vikas supplied details that led Delhi Police to identify and arrest the other four suspects. Police say each accused had a specific role, from technical support for the voice‑cloning software to managing the financial transfers.
Mumbai Police had already registered a cyber‑fraud case after the company lodged a formal complaint. As the scale of the scheme became clearer, Delhi Police joined the investigation. The two agencies are now collaborating to trace the full financial trail, identify additional beneficiaries, and determine whether the suspects belong to a larger cyber‑crime network operating across state lines.
Voice‑cloning, also known as deepfake audio, uses AI models to generate speech that mimics a target individual. The technology can produce realistic voice outputs from only a few seconds of audio. While originally developed for entertainment and accessibility applications, it has become increasingly accessible to the public and can be misused for social‑engineering attacks.
Cybersecurity experts note that such scams rely on manipulating human decision‑making rather than hacking systems directly. By convincing employees that a senior executive is issuing orders, attackers bypass technical controls. Experts advise organisations to adopt stronger safeguards, such as multi‑factor authentication for financial approvals, call‑back verification procedures, and employee training on deepfake awareness.
The case underscores the growing threat of AI‑driven financial fraud in India. According to reports, the use of synthetic media for scams has risen sharply in 2026, with several incidents involving voice‑cloning and deepfake video. The Delhi‑Mumbai scam is one of the first high‑profile cases where AI was used to orchestrate a large‑scale corporate fraud.
At present, the five suspects remain in custody, and Delhi and Mumbai Police are continuing to investigate the possibility of a wider network. The company that was defrauded has not released a statement, but industry observers expect it to review its internal controls and consider implementing voice‑authentication technologies.
The incident highlights the need for coordinated law‑enforcement efforts to tackle technologically sophisticated crimes that cross state borders. It also raises questions about regulatory frameworks for AI tools that can be weaponised for fraud.
As of 19 June 2026, the investigation is ongoing. No further arrests have been announced, and the company has not confirmed the recovery of the stolen funds. The case remains a cautionary example of how AI can be exploited to undermine trust and financial integrity in corporate settings.
