We advise both property clients and technology businesses, and one question now comes up regularly from the second group: if a PropTech platform uses AI to generate listing descriptions, answer buyer chatbot queries, or produce automated property valuations, and that output turns out to be wrong, who is legally responsible? The honest answer is that Indian law has not yet built a bespoke framework for this, so liability is worked out by applying existing law, contract, tort, consumer protection, and data protection, to a new kind of output.
Start With What the Platform Actually Represents
The first question in any dispute is not "was this AI-generated" but "was this a representation the buyer relied on". If an AI-generated listing states a flat has clear title, a specific carpet area, or that RERA registration is in place, and that turns out to be false, the platform can face liability for misrepresentation regardless of whether a human or a model wrote the sentence. Courts and consumer forums look at reliance and harm, not authorship.
Intermediary Safe Harbour Has Limits
Platforms often assume Section 79 of the Information Technology Act, the "intermediary" safe harbour, protects them from liability for third-party content. That protection is conditional, and it weakens considerably once the platform's own AI system generates or materially edits content, rather than merely hosting what a user submitted. A platform that auto-generates a property description from raw data it holds is closer to being a publisher of that content than a passive intermediary, and safe harbour arguments become harder to sustain.
Where the DPDP Act Comes In
Separately from content liability, if the AI system is trained on or processes personal data, buyer contact details, seller identity, financial information, it triggers obligations under the Digital Personal Data Protection Act, 2023. A PropTech company acting as a Data Fiduciary has to justify the purpose of processing, secure consent where required, and be able to explain how personal data feeds into automated outputs shown to users. We covered the penalty exposure for getting this wrong in our piece on DPDP Act penalties.
Automated Valuations Carry Their Own Risk
AI-generated property valuations sit in a particularly exposed position. If a platform's valuation tool materially overstates a property's worth and a buyer relies on it to justify an offer, or a lender relies on it indirectly, questions of professional negligence and deficiency of service under the Consumer Protection Act can arise, even where the platform's terms of service disclaim the valuation as "indicative only". Disclaimers reduce risk, they don't eliminate it, particularly where the platform's marketing elsewhere describes the tool as accurate or expert-backed.
Practical Steps We Recommend to PropTech Clients
- Maintain a clear internal record of what data trains or informs each AI output, so you can explain and defend it if challenged.
- Build a human review step for any AI output that makes a factual legal claim, RERA status, title status, area figures, rather than publishing model output directly.
- Draft terms of service and disclaimers specific to AI-generated content, not recycled from a general liability clause, and have them reviewed against how the tool is actually marketed.
- Treat your AI governance obligations and your DPDP compliance as connected projects, not separate checkboxes, since the same system usually implicates both.
This is an evolving area, and enforcement patterns in India are still forming. If your platform generates property content, valuations, or buyer-facing automated advice, it is worth getting a legal opinion on your specific exposure before, not after, a dispute arises. See our AI Liability & Risk Advisory service or contact us to discuss your platform's specific setup.