Artificial intelligence is no longer a future technology. It is already influencing how Indians work, learn, shop, invest, access healthcare, and interact with public services. Businesses are using AI to improve customer support, automate workflows, analyse data, and create content. Governments are exploring AI for citizen services, fraud detection, and administrative efficiency. At the same time, citizens are increasingly concerned about misinformation, privacy, bias, job displacement, and the misuse of personal data.
This creates a central policy question for India: how can the country regulate AI without slowing down the innovation it needs? The answer is not to choose between unrestricted experimentation and heavy-handed control. India needs guardrails that empower responsible innovation. It needs rules that protect people from real harms while giving entrepreneurs, enterprises, researchers, and public institutions the confidence to build with AI.
This balanced view is reflected in Jaspreet Bindra’s perspective on AI regulation in India. His argument is practical: regulate known risks first, build AI literacy at scale, and avoid an overly broad law that may become outdated before it can be effectively implemented. This approach recognises that AI is evolving quickly and that regulation must be agile enough to keep pace.
India has already shown that it can build digital systems at population scale. Aadhaar simplified identity verification for more than a billion people. UPI transformed digital payments and made financial transactions accessible to millions who were previously outside the formal economy. These successes were not accidental. They were built through a combination of public infrastructure, private innovation, policy support, and continuous iteration. AI can follow a similar path if India treats it as a national capability rather than merely another technology trend.
The first principle of effective AI regulation should be proportionality. Not every AI application carries the same level of risk. A customer-service chatbot, a content-generation tool, and an AI-powered recommendation engine may require basic transparency and data-protection safeguards. But systems that influence employment, credit, healthcare, policing, education, or access to essential services require stronger oversight. A one-size-fits-all law could burden low-risk innovation while failing to address the most serious harms.
India should therefore adopt a risk-based framework. Low-risk applications may need clear disclosure, privacy protections, and basic accountability. High-risk applications should require human review, bias testing, audit trails, and accessible mechanisms for redress. This approach protects citizens without placing unnecessary burdens on every business experimenting with AI. It also gives companies a clearer path to compliance.
The second principle should be targeted regulation of visible harms. Deepfakes are one of the most urgent examples. They can impersonate public figures, manipulate voters, defraud individuals, harass women, and spread misinformation during sensitive events. Existing laws may offer some remedies, but they are often reactive. By the time a fake video is removed, the damage may already have been done. India needs faster takedown mechanisms, clear labelling requirements for synthetic media, stronger penalties for malicious impersonation, and better support for victims.
Voice cloning is another growing concern. Criminals can use AI to imitate the voice of a senior executive and instruct employees to transfer money or share confidential information. They can impersonate family members in emergency scams. They can create fake customer-support videos to trick people into sharing banking credentials. These are not abstract risks. They are direct threats to financial security and personal safety.
Copyright is another area that requires practical attention. Generative AI systems are trained on enormous volumes of human-created content, including books, articles, images, music, code, and videos. Creators deserve recognition and fair compensation when their work contributes to commercial AI systems. At the same time, India should not create a system so restrictive that it prevents domestic AI development. Revenue-sharing arrangements, licensing frameworks, citation norms, and transparent data-use practices can create a more sustainable middle ground.
The third principle should be AI literacy. Regulation can define what companies must do, but it cannot teach every employee how to use AI responsibly. AI literacy is different from technical training. A marketing executive does not need to build a machine-learning model, but they should understand how to verify AI-generated claims, protect confidential information, identify bias, and recognise when human judgement is required. A teacher should know how to use AI to support learning without allowing it to replace critical thinking.
India should embed AI literacy in schools, colleges, professional education, and workplace training. The objective should not be to turn every citizen into an AI engineer. Instead, it should be to help people become confident, critical, and responsible users of AI. This matters because the technology will keep changing. Tools that are popular today may be replaced tomorrow, but the ability to question outputs, understand limitations, and apply ethical judgement will remain valuable.
Businesses also have a responsibility to act before regulation becomes mandatory. Companies should create internal AI policies covering data privacy, human oversight, vendor selection, employee training, and acceptable use. They should identify high-risk use cases, such as automated hiring, credit decisions, healthcare support, and customer complaints, and apply stronger controls in those areas. Responsible adoption is not only about avoiding penalties. It is about building products and services that customers can trust.
Trust will become a major competitive advantage in the AI era. Consumers may tolerate imperfect technology, but they are less likely to tolerate hidden data practices, discriminatory outcomes, or unaccountable decisions. Companies that invest in transparency, privacy, and human review will be better positioned to earn long-term customer loyalty. In this sense, responsible AI is not a cost centre; it is a business capability.
India’s policymakers should avoid the temptation to copy regulatory models from other countries without adaptation. The European Union has developed a comprehensive AI framework, while the United States has relied more heavily on market-led innovation and sector-specific oversight. Both approaches offer lessons, but neither is perfectly suited to India’s needs. India should create a model that reflects its own digital infrastructure, entrepreneurial energy, linguistic diversity, and development priorities.
The country also needs regulatory capacity. Rules are only useful if regulators understand the technology and can respond to emerging risks. This requires technical expertise, inter-agency coordination, data access, and a willingness to update rules as the technology evolves. A static law written for today’s AI may not be suitable for the systems that emerge three years from now. India should therefore favour agile, principle-based governance that can evolve with the technology.
Civil society, academia, industry, and government must work together. Public consultations, regulatory sandboxes, industry standards, and independent audits can help create rules that are practical rather than theoretical. India should invite startups, large enterprises, researchers, creators, and consumer groups into the conversation. AI policy should not be written only by technology companies or only by regulators. It should reflect the interests of everyone who will be affected by it.
The ultimate goal is not to slow AI down. It is to make AI adoption safer, fairer, and more widely beneficial. India has the talent, digital infrastructure, and market scale to become a global AI leader. What it needs now is a governance model that matches that ambition. By focusing on real harms, building AI literacy, protecting creators, and enabling responsible innovation, India can create an AI ecosystem that serves both economic growth and public trust.
A balanced approach will require patience, coordination, and political will. But the alternative—either ignoring AI or regulating it out of fear—would be worse. India should choose a third path: innovation with guardrails, progress with accountability, and technology that works for people. This is the approach that can help India turn AI from a source of anxiety into a source of national advantage.
Author bio:AI & Beyond is a platform dedicated to building AI literacy and helping individuals and organisations understand, adopt, and govern artificial intelligence responsibly. Its insights explore AI strategy, generative AI, digital transformation, ethics, regulation, and the future of work. Explore the original discussion on Jaspreet Bindra on AI regulation in India.