Indian Financial Firms Face Rising AI Compliance Risks Before New Law

Indian financial institutions face increasing legal and regulatory scrutiny over artificial intelligence use, even as comprehensive national legislation remains pending. Recent court rulings and central bank directives indicate that existing laws already hold companies accountable for AI-driven errors and system failures.

Key Takeaways

• Accountability remains with the user: Financial institutions are responsible for risks associated with AI models, including those licensed from third-party vendors.
• Regulatory action precedes legislation: The Reserve Bank of India is actively enforcing risk management frameworks while national AI law is still in development.
• Judicial intervention is rising: Courts are overturning decisions based on flawed AI-generated data, emphasizing that companies cannot rely on automation to bypass legal scrutiny.
• Board-level oversight is required: Regulators are demanding direct board-level engagement on cyber risk assessments rather than delegating responsibility to technical staff.
• Existing laws apply: New statutes will complement, not replace, current case law and the upcoming Digital Personal Data Protection Act.

Judicial Scrutiny of AI Citations

The Supreme Court of India recently intervened in Pooja Ramesh Singh v. Jammu and Kashmir Bank Ltd. on July 2, 2026. The court quashed orders from the National Company Law Tribunal and the National Company Law Appellate Tribunal that relied on AI-generated citations incorrectly attributed to the Supreme Court. The ruling prompted the Bar Council of India to establish disciplinary norms for legal professionals submitting AI-generated content.

Legal analysts suggest this case serves as a warning for businesses that assume AI workflows reduce liability. When automated systems, such as those used by banks for transaction monitoring, produce errors that human supervisors fail to catch, the institution remains liable for the resulting legal and financial consequences. Courts are applying existing legal principles to these scenarios rather than waiting for specific AI statutes.