Introduction
In December 2025, India moved beyond guidelines into hard law. The Artificial Intelligence (Ethics and Accountability) Bill, 2025, tabled by MP Bharti Pardhi, represents India’s first statutory framework to regulate Artificial Intelligence (“AI”) related ethics. It introduces a statutory Ethics Committee for AI, imposes fines up to Rs. 5 crore, and sets strict rules on surveillance and algorithmic decision-making. For lawyers, technologists, and business leaders, this bill demands immediate attention.
The need for such a bill in India
The previous approach where only guidelines were issued proved insufficient. Three critical gaps needed addressing:
- Algorithmic bias at scale: AI systems inherit and amplify discrimination. When decisions on loans, jobs, or law enforcement are automated, bias affects vulnerable communities disproportionately.
- Unaccountable surveillance: Facial recognition, tracking systems, and lack of transparency in decision making, behaviour prediction tools without oversight.
The Three Pillar Framework
-
- Establishment of the Ethics Committee for Artificial Intelligence: The AI (Ethics and Accountability) Bill, 2025 establishes a statutory Ethics Committee for AI as its central regulatory mechanism. Unlike advisory bodies, the committee is vested with substantive powers to frame ethical guidelines, monitor compliance, and investigate instances of AI related harm. The committee is structured as a multi stakeholder body comprising experts in ethics, technology, law, data science, and human rights, along with representatives from academia, industry, civil society, and government. This composition ensures that AI governance is informed by technical expertise as well as legal and social considerations. Importantly, the committee has investigative authority to examine complaints of algorithmic bias or misuse and to recommend remedial measures or penalties. This marks a shift towards institutionalised accountability for ethical breaches in AI development and deployment.
- Regulation of high risk AI applications: The Bill adopts a risk based approach by imposing strict controls on high risk AI uses. AI driven surveillance is permitted only for lawful purposes and requires prior approval from the Ethics Committee, introducing preventive ethical oversight before deployment. In critical decision making areas such as law enforcement, credit assessment, and employment, the Bill prohibits discrimination on grounds of race, religion, or gender and mandates rigorous ethical review. These safeguards extend constitutional principles of equality and non-discrimination under Articles 14, 15, and 16 of the Indian Constitution to algorithmic decision making.
- Accountability and Transparency: The Bill places clear obligations on AI developers to ensure transparency and accountability. Developers must disclose the purpose, limitations, training data, and decision making logic of AI systems, effectively curbing opaque “black box” models. Regular bias audits and inclusive training datasets are mandated, and systems exhibiting significant bias must be withdrawn until corrected. Developers are also required to maintain detailed compliance records, creating enforceable audit trails.
What This Means for You?
- For Lawyers: AI is automating legal work document review, contract analysis, case prediction. The ethics bill does not ban these tools but imposes requirements on how one can use them. If a firm uses AI, it must verify that AI was developed with transparency and bias audits are in place. The firm is responsible for understanding how it works and ensuring it does not harm clients.
- For AI Developers: The bill creates new compliance obligations. Deploying sensitive AI systems now requires Ethics Committee engagement. Therefore, companies that build ethical safeguards into products from day one will have competitive advantages. Those treating compliance as an afterthought will struggle.
- For Businesses Using AI: Finance, e-commerce, healthcare, Human resource, all such areas that use AI in operations will need to audit systems for bias, understand how they work, and be ready to justify their use to the Ethics Committee.
Enforcement: Real Consequences
Violations result in:
- Financial penalties: Up to ₹5 crore, scaled by severity.
- License suspension or revocation: Companies cannot deploy AI systems if it violates its obligations under the bill.
- Criminal liability: Repeat or serious violations can attract criminal charges, not just fines.
How This Fits India’s AI Ecosystem?
The bill does not operate alone. It is part of a broader framework:
- India AI Governance Guidelines of November 2025: The guidelines lay down seven core ethical principles: trust, people first approach, innovation, fairness, accountability, understandability, and safety. The bill operationalises these through a legislative framework.
- AI Safety Institute (AISI): Provides technical expertise on AI risks and safety testing
- Digital Personal Data Protection Act, 2023: This act and the rules therein govern how personal data which is also used for critical training AI is collected and protected.
Hence, a multi-layered framework will be in place in which the guidelines provide principles, the bill provides legal enforceability, regulators provide specific rules, and the Digital Personal Data Protection Act, 2023 protects the data used in AI training and functioning.
India’s move to codify AI ethics into law reflects a global realisation that AI is very important, and its risks are too significant to rely on self-regulation and guidelines alone. The EU went this route with the AI Act. China has regulations on algorithmic recommendation systems. The US is moving toward sector specific AI governance. India’s approach of establishing an ethics based review process, and imposing penalties including criminal liability is a comprehensive structure. This is about ensuring that the benefits of AI that is innovation, efficiency, and progress do not come at the cost of democratic values, individual rights, and social equity.
Conclusion
The Artificial Intelligence (Ethics and Accountability) Bill, 2025 represents a watershed moment for AI governance in India. It moves the conversation from abstract principles to concrete legal obligations. It empowers an institutional authority to enforce ethical standards. The bill ensures that the development as well as the use of AI in India will become more accountable, responsible and also prevent misuse that can lead to infringement of rights.
