NEW DELHI, October 24: The Indian government is preparing to move towards a dedicated legal framework for artificial intelligence, signalling a shift in its approach to regulating a technology that is rapidly transforming businesses, public services and digital communication.
Union Electronics and Information Technology Minister Ashwini Vaishnaw said on October 23 that the government would release a consultation paper on a proposed artificial intelligence law in the coming weeks. He argued that the risks associated with AI had reached a point where legislation was necessary to address potential harm while ensuring that innovation continued.
The announcement marks an important development in India’s AI policy, as authorities seek to establish safeguards without imposing restrictions that could slow the country’s growing technology ecosystem.
The proposed framework is expected to examine how AI-related risks can be addressed while maintaining human oversight over increasingly capable systems. The government has not yet announced a final legislative proposal, a timetable for introducing a bill in Parliament or details of the regulatory structure that may eventually be established. Government signals a shift towards legislation
India’s approach to AI regulation has so far emphasised the use of existing laws, targeted amendments and governance guidelines rather than an entirely separate statute.
The India AI Governance Guidelines released in November 2025 favoured adapting existing legal provisions instead of immediately creating a standalone AI law. The government’s latest position suggests that it is reconsidering whether those arrangements will be sufficient as the technology becomes more powerful and widely deployed.
Speaking at the NDTV World Summit in New Delhi, Vaishnaw said the government was working on a new regulatory approach to address harmful applications of AI.
He indicated that the objective would be to regulate the damage the technology could cause rather than prevent its development. The minister also stressed that innovation and public safety would have to advance together.
The consultation paper will provide an opportunity to gather views before the government settles on the design of the proposed framework. However, the precise obligations that may apply to developers, companies and organisations using AI remain undecided.Human control at the centre of the proposed framework
One of the main concerns raised by the government is ensuring that humans retain meaningful control over AI systems and the actions they perform.
Modern AI applications increasingly extend beyond generating text or images. AI agents can interact with software, retrieve information, execute multi-step tasks and make decisions within systems connected to business operations.
These capabilities offer opportunities to improve efficiency, but they can also create problems when automated systems act outside their intended limits or access information they were not authorised to use.
Vaishnaw has argued that control over AI models, agents, processes and outcomes must remain in human hands. He also referred to the need for mechanisms that can stop systems when their behaviour moves beyond acceptable boundaries. <escape>【turn651315search0】</escape>
Such principles could become important considerations in the consultation process. Policymakers may need to examine how human supervision should work in practice, including when an automated system makes a consequential decision or performs actions without direct approval.
A practical framework would also need to distinguish between low-risk applications, such as routine administrative assistance, and higher-risk uses involving financial transactions, sensitive personal data or critical infrastructure.
The government has not yet specified how these distinctions would be defined or enforced.
Deepfakes and misuse of synthetic content
The spread of AI-generated images, videos, audio and written material has increased concerns about misleading content circulating online.
Deepfake technology can create realistic representations of people saying or doing things that never occurred. Such material can be used for entertainment and legitimate creative work, but it can also facilitate impersonation, fraud, harassment and the spread of false information.
Earlier statements by the government indicated that AI regulation would consider deepfakes, cybersecurity, safety and the distribution of benefits from the technology.A future legal framework may need to consider how to distinguish legitimate synthetic content from material intended to deceive or cause harm. It may also need to address disclosure requirements, complaint mechanisms and the responsibilities of platforms distributing misleading content.
However, the details of any such requirements have not been finalised. Questions remain about how rules would apply to different forms of content, what exceptions might be necessary and how authorities could respond quickly when harmful material spreads across multiple services.
The challenge is to protect people from deception without unnecessarily restricting satire, artistic expression, research or other legitimate uses of generative AI.
Balancing regulation with technological growth
The government has indicated that it does not want regulation to become an obstacle to technological progress.
Countries around the world are adopting different approaches to AI governance. Some favour detailed legal obligations and stronger controls over high-risk systems, while others place greater emphasis on innovation, voluntary standards and targeted interventions.
Vaishnaw has highlighted the need to avoid an approach that restricts new technology too early, while maintaining safeguards for citizens. He has also argued that AI governance should be internationally aligned because digital systems routinely operate across national borders.
For India, the balance is especially significant as companies, start-ups, research institutions and public agencies expand their use of AI.
Excessively complicated requirements could raise compliance costs for smaller firms and discourage experimentation. On the other hand, insufficient safeguards could expose consumers and organisations to fraud, privacy violations, unreliable automated decisions and security failures.
A carefully designed framework would need to provide clear expectations for businesses while allowing rules to evolve as the technology changes.
Consultation to include users of AI
The proposed consultation is expected to extend beyond technology developers to include civil society, citizens, industry representatives and organisations that use AI in their operations.
This broader participation could help policymakers understand how AI affects people outside the technology sector.
Banks, insurance companies, manufacturers and service providers are among the organisations increasingly exploring AI applications. Their experiences could help identify practical risks associated with automated decisions, sensitive information and the reliability of AI-generated outputs.
For example, an organisation using AI to assist with customer service may need different safeguards from a financial institution using automated systems to assess applications or identify suspicious transactions.
Consultation with users could also help reveal problems that may not be apparent during the development of an AI model but emerge when it is deployed in a real-world environment.
The government has not yet published the final consultation document, so the questions it will ask and the period allowed for responses remain to be confirmed.
Cybersecurity concerns add urgency
The debate over regulation comes amid growing concerns about AI agents behaving unexpectedly and the security implications of increasingly capable systems.
On October 23, OpenAI announced a research partnership with Poland’s NASK National Research Institute to examine AI safety, security and technical risks. The collaboration will focus on evaluating advanced systems, identifying vulnerabilities and improving safeguards.
The initiative reflects wider efforts to establish more reliable methods for testing AI systems before and during deployment.
In India, the government’s concerns also intersect with the challenge of ensuring that AI applications do not create new cybersecurity weaknesses. Automated agents may require access to files, online services and internal databases to complete tasks. If their permissions are poorly designed, they could expose confidential information or perform unauthorised actions.
Regulation alone cannot eliminate these risks. Organisations will also need appropriate technical controls, regular testing, monitoring and clear accountability for the use of automated systems.
The proposed law could provide a broader policy framework, while detailed technical standards and sector-specific safeguards may be required to address different applications.
Implications for businesses and start-ups
The direction of India’s AI policy will matter to companies deciding how to develop and deploy new products.
Clear rules could help businesses understand their responsibilities and build greater trust among customers. Common standards for safety testing, transparency and accountability may also make it easier for organisations to adopt AI in sensitive areas.
At the same time, businesses will need clarity on the costs of compliance, the treatment of open-source models, liability for AI-generated harm and the responsibilities of companies that integrate third-party AI tools into their services.
Smaller enterprises may require practical guidance and accessible compliance measures to avoid being disproportionately burdened by complex regulations.
Start-ups and research institutions will also be watching how the proposed framework addresses experimentation and innovation. A flexible approach could support development while introducing stronger protections where the potential for harm is greater.
The government has not yet provided enough detail to determine how these issues will be resolved.
International coordination remains important
AI systems are developed, trained and deployed across borders, often using international computing infrastructure and data services. As a result, national regulations may be difficult to enforce effectively without cooperation between governments and technology companies.
Vaishnaw has emphasised the importance of global alignment, suggesting that countries should work towards common principles for managing AI-related risks. International cooperation could help establish comparable methods for evaluating advanced systems, sharing information about serious incidents and improving safeguards against misuse.
However, countries may differ over privacy, freedom of expression, competition, national security and the degree of control that should be exercised over AI developers.
India will therefore have to consider international practices while designing rules suited to its own legal system, digital infrastructure and economic priorities.
What happens next?
The forthcoming consultation paper is expected to begin a more formal discussion about the structure and scope of AI regulation in India.
Its publication should provide greater clarity on the issues the government wants to address and the options being considered. Responses from industry, civil society, researchers and citizens could then inform the development of a legislative proposal.
Important questions remain open, including whether the law will establish a dedicated regulator, how compliance will be assessed, what penalties may apply and how responsibilities will be divided between AI developers and organisations using their products.
The government will also need to consider how quickly the framework can adapt as AI capabilities evolve.
India’s challenge is to create rules that protect citizens without preventing beneficial applications from reaching the public. The success of the proposed framework will depend not only on the legislation itself but also on clear implementation, technical expertise and effective cooperation between regulators and the technology sector.
The planned consultation represents an initial step towards defining that balance. Until the paper is released and the government provides further details, the precise shape of India’s proposed AI law remains uncertain.