Millions of listings need mandatory declarations in India, so marketplaces use AI agents to check maximum retail price and origin fields, test journeys for dark patterns, confirm seller details and review consent and children's data handling. Compliance keeps every takedown and regulator response. Consumer protection rules, RBI's payment rules and the DPDP Act 2023 apply.
What fills e-commerce compliance work in India?
Listing declarations are the daily grind. Packaged goods sold online must show mandatory declarations on the listing, including maximum retail price, net quantity, manufacturer or importer details and country of origin. Marketplaces with millions of listings from thousands of sellers must make sure those fields are present and accurate.
Marketplace duties follow. E-commerce consumer protection rules require marketplaces to show seller details, return and refund terms, and a grievance officer, and to avoid manipulating prices. India's consumer protection authority has issued guidance against dark patterns, covering practices such as false urgency, items sneaked into the basket, drip pricing and subscriptions that are hard to cancel. Compliance reviews journeys against that guidance.
Payments and data complete the load. RBI rules stop merchants storing customers' card numbers, so saved cards depend on tokens. The DPDP Act requires notice and consent for processing personal data, with extra care for children's data. Foreign-owned marketplaces must also stay within the rules that separate a marketplace from an inventory-holding retailer.
Where can AI agents help?
| Use | What the agent does | Depth |
|---|---|---|
| Declaration checks | Checks each packaged-goods listing for maximum retail price, net quantity, manufacturer or importer and country of origin, and drafts fix requests to sellers | 2 |
| Dark pattern review | Walks journeys on web and app and flags countdowns that reset, pre-ticked add-ons, late fees and hard-to-find cancellation | 1 |
| Seller detail display | Confirms each seller's name, address and contact are displayed and match onboarding records | 1 |
| Consent and children's data | Checks that consent notices match actual processing and flags journeys where children may be users without the required parental consent | 1 |
| Grievance triage | Classifies complaints in English, Hindi and regional languages and drafts responses for the grievance officer | 2 |
Declaration checks are where scale makes an agent essential. A human team cannot check millions of listings; an agent can check every new listing on arrival and the whole catalogue on a cycle, and send sellers standard fix requests through a person.
What stays with people?
Delisting products, suspending sellers and changing marketplace policy stay with named staff. These affect sellers' livelihoods and the marketplace's legal position. The agent drafts requests and recommendations; people decide.
The grievance officer's responses stay with that officer, who is named under the rules. Responses to the consumer protection authority, and any change to how the business separates marketplace and inventory activity, stay with compliance and legal.
Judgements about dark patterns can be fine. Whether a stock message is honest or creates false urgency depends on the data behind it. The agent flags; a person confirms. This is the idea in AI compliance agents: second line, never the last word.
Which rules apply in India?
India's consumer protection law and its e-commerce rules set duties for marketplaces and sellers, including disclosure, grievance handling and fair pricing, and the consumer protection authority has issued guidance on dark patterns. Packaged goods rules require mandatory declarations on listings. Foreign investment rules distinguish the marketplace model from the inventory model. RBI regulates payments, including the rule that merchants may not store card numbers.
The Digital Personal Data Protection Act 2023 governs customers' personal data. The business is the data fiduciary: it must give notice, obtain consent for consent-based processing, honour withdrawal of consent and other rights, secure data, and obtain verifiable parental consent before processing children's data. See how AI agents are regulated in India for the wider framework.
When customers use AI agents to shop for them, disclosure duties still apply to what those agents read. Our explainer on what agentic commerce is covers the shift.
This is an explanation of the rules, not legal advice; confirm how they apply to you with counsel.
How should you start in the first 90 days?
Weeks one to four: run declaration checks at depth 1 on one high-volume category, such as packaged food or personal care. Measure how many listings lack declarations and which sellers account for most gaps.
Weeks five to eight: run the dark pattern review across your top journeys on web and app, and fix confirmed issues. Record each change.
Weeks nine to twelve: move declaration checks to depth 2 with fix requests to sellers held for a person, and start grievance triage in your main languages.
Where Pryme fits
Pryme certifies declaration checks and grievance triage separately, each on the marketplace's own listings and complaints, through Attest. Seller suspension is outside the checking agent's scope and is refused. Every verdict is filed, so the marketplace can show the consumer protection authority what it checked and when.
Questions people ask
Can an AI agent delist a seller's products automatically?
It can, but we recommend holding delisting for a person except in clear cases, such as a banned product. Wrong delistings harm sellers and may breach the marketplace's own terms.
Does the DPDP Act apply to a shopping assistant chatbot?
Yes, if it processes personal data of customers in India. Notice, consent, purpose limitation and security apply, as for any other processing.
How does India compare with Germany for e-commerce compliance?
Germany adds private enforcement through warning letters and strict checkout wording rules. See our brief on compliance agents in German e-commerce.