AI Safety in India

Deepfakes, Privacy Risks and Child Protection in the Digital

Introduction

Artificial intelligence (AI) is transforming how people work, study, communicate and access information in India. From AI-powered chatbots and digital assistants to image generators and automated business tools, the technology offers numerous opportunities for individuals and organisations. However, its rapid growth has also raised serious concerns about deepfakes, privacy violations, online fraud, misinformation and children’s safety.

On October 14, 2026, discussions at the Trust and Safety India Festival in New Delhi highlighted the growing challenge of keeping people safe as AI technology develops. Experts examined deepfakes, child protection, privacy concerns and the need for stronger safety measures. The discussion reflects an important question for India’s digital future: how can the country encourage AI innovation while protecting its citizens?

What Is AI Safety and Why Does It Matter?

AI safety refers to the measures used to ensure that artificial intelligence systems operate responsibly and do not cause unnecessary harm. These measures include protecting personal information, preventing harmful content, reducing discrimination, improving transparency and making AI systems more reliable.

AI tools can help students understand difficult subjects, assist doctors with information, improve business productivity and support people with disabilities. At the same time, poorly designed or deliberately misused systems can spread false information, expose private data and enable criminals to impersonate others.

AI safety is therefore not only a technical issue. It also involves education, responsible business practices, government oversight and public awareness.

1. Deepfakes: A Growing Digital Threat

Deepfakes are images, videos or audio recordings created or altered using AI to make someone appear to say or do something they never actually said or did. Modern tools can imitate a person’s face, voice and expressions with increasingly convincing results.

One major concern is financial fraud. A criminal might use a cloned voice to impersonate a family member, company executive or bank representative and ask for an urgent money transfer. Similarly, a manipulated video featuring a public figure could spread false claims before people have an opportunity to verify the information.

Deepfakes can also damage reputations, manipulate public opinion and enable harassment. Women and young people can be particularly vulnerable to the misuse of their photographs in fabricated, sexually explicit content.

The Indian government has strengthened its regulatory framework for AI-generated content. In March 2026, the Ministry of Electronics and Information Technology described measures addressing harmful synthetic content, including deepfakes, impersonation and material that threatens children’s safety. The framework includes platform responsibilities and requirements concerning the identification and handling of permitted synthetic content.

The government has also indicated that further AI regulation is being considered, with safety and deepfakes among the priorities announced in October 2026.

2. Online Privacy Risks in the AI Era

AI systems often rely on information supplied by users or collected through digital services. This can include names, photographs, voice recordings, location information, browsing activity and other personal details.

Sharing too much information with unfamiliar applications can create privacy risks. For example, uploading identity documents, financial records or confidential business information to an untrusted AI tool may expose information that should remain private.

AI-powered impersonation can make these risks worse. Public photographs and voice recordings may be misused to create fake accounts or convincing fraudulent messages.

Users can take several steps to protect their privacy:

• Avoid entering passwords, banking details or confidential documents into untrusted AI tools.
• Review application permissions before granting access to contacts, photographs, microphones or location information.
• Use strong, unique passwords and enable two-factor authentication wherever available.
• Avoid sharing identity documents or private photographs unnecessarily.
• Read privacy policies and understand how services may store or use uploaded information.
• Verify unexpected requests for money or sensitive information through a separate, trusted communication channel.

Businesses should also limit employee access to confidential data, establish clear rules for AI usage and review the security practices of third-party tools.

3. Protecting Children From AI-Related Harm

Children are increasingly exposed to digital platforms for education, entertainment and communication. AI can provide personalised learning experiences and creative opportunities, but it can also expose children to harmful content, cyberbullying, inappropriate interactions and misleading information.

One serious risk is the creation or distribution of manipulated images involving children. AI-generated material can be used to harass, exploit or intimidate young people. Children may also struggle to recognise when an online person, photograph or voice message is fake.

Parents and guardians can help by teaching children not to share private photographs, home addresses, school details or passwords with strangers. They should also encourage children to report uncomfortable online interactions without fear of punishment.

Useful precautions include:

1. Set age-appropriate privacy and parental-control settings.
2. Discuss the difference between genuine and AI-generated content.
3. Encourage children to verify information before believing or forwarding it.
4. Keep communication open so children can report suspicious accounts or messages.
5. Teach children never to meet an online contact without appropriate adult supervision.
6. Report suspected exploitation, threats and harmful content to the relevant platform and authorities.

Schools can support these efforts through digital-literacy lessons that explain online privacy, responsible AI use, misinformation and cyberbullying.

4. India’s Legal and Regulatory Response

India has existing laws and rules that address several forms of digital harm, even as policymakers consider how AI-specific regulation should develop.

The Information Technology Act, 2000, contains provisions addressing offences such as identity theft, online impersonation and certain privacy violations. The Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, as amended, establish responsibilities for digital intermediaries concerning unlawful content and harmful synthetic media.

The Digital Personal Data Protection Act, 2023, is also part of India’s framework for protecting personal data, subject to its applicable provisions and implementation.

In March 2026, the government reported amendments to the IT Rules aimed at addressing harmful AI-generated content, including impersonation, misleading material and content that harms children. The measures also address the labelling and traceability of permissible synthetic content and faster action on unlawful material under applicable directions.

However, laws alone cannot eliminate every risk. Effective enforcement, accessible reporting mechanisms, cooperation between platforms and authorities, and public education remain essential.

5. How to Identify and Report AI-Generated Content

It is becoming harder to identify deepfakes simply by looking at a photograph or listening to a recording. Some manipulated content may contain visual errors, unusual expressions or unnatural audio, but advanced tools can produce highly convincing results.

People should therefore focus on verification rather than relying entirely on visual clues.

Before sharing a shocking video or recording, check whether credible news organisations or official sources have independently confirmed it. Look for the original source, publication date and context. If someone requests urgent payment through a voice message, call them using a previously verified phone number.

If you encounter harmful or deceptive content, preserve relevant evidence where safe and lawful, report the content through the platform’s reporting system and contact the appropriate authorities when necessary. In India, cybercrime complaints can be submitted through the National Cyber Crime Reporting Portal at cybercrime.gov.in. Suspected financial cyber fraud can also be reported through the national helpline 1930.

Do not redistribute intimate or exploitative material as a way of warning others, particularly when children may be involved.

6. The Future of Responsible AI in India

India has an opportunity to become a leader in responsible AI development. Achieving this goal requires collaboration among technology companies, researchers, policymakers, educators, parents and ordinary users.

AI developers should test systems for foreseeable misuse, improve privacy protections and establish clear ways to report harmful outputs. Social media platforms should improve their detection systems while providing transparent and accessible complaint procedures. Businesses should train employees to recognise AI-enabled fraud and avoid exposing sensitive information.

Governments must continue to assess whether existing laws and new regulatory measures provide adequate protection without unnecessarily restricting beneficial innovation.

Most importantly, AI literacy should become a regular part of education. People need to understand that convincing digital content is not necessarily genuine and that even familiar voices and faces can be manipulated.

Conclusion

AI safety in India is becoming increasingly important as artificial intelligence enters everyday life. Deepfakes can enable fraud and misinformation, careless data sharing can expose private information, and children face particular risks from online exploitation and harmful content.

Recent discussions in India demonstrate the growing importance of stronger safeguards, responsible technology development and effective public awareness. Government regulation is an important part of the response, but technology companies, schools, parents and individual users also have essential roles to play.

The goal should not be to stop AI innovation. Instead, India should encourage its benefits while ensuring that digital technology respects privacy, protects children and supports a safer online environment for everyone.

Responsible AI begins with informed users, secure technology and accountability.