Article

Nov 4, 2026

India’s AI Rules 2026: 5 Things Small Businesses Should Do Now

India’s AI consultation paper is due within weeks. See which AI rules already apply and 5 steps small businesses can take now.

Black blog header with orange glow and the title India’s AI Rules 2026: 5 Things to Do Now

By Xvox Labs Team · Last updated: 4 November 2026. Researched with AI tools and reviewed by the Xvox Labs team.

Quick answer: India’s IT Minister says an AI regulation consultation paper is due “in the coming weeks”; details aren’t public. Platforms must already label AI-made images, audio and video (since 20 February 2026), and data protection duties phase in by May 2027. Small businesses using chatbots, AI visuals or customer data should act now: five steps work whatever the draft says.

This post is general information, not legal advice. For decisions about your business, speak to a lawyer.

What AI rules is India planning in 2026?

India’s IT ministry (MeitY) is preparing a consultation paper on AI regulation, which is a draft for public comment, not a final law. IT Minister Ashwini Vaishnaw first said it would come “in a month’s time” (8 October 2026), then “in the coming weeks” (26 October). He has also said a new law may be needed; the contents are not public yet.

What has been said on the record:

  • 8 October 2026: built on safety, a human-first approach, skilling and no one left behind, plus deepfake harms (ThePrint).

  • 26 October 2026: He said “We will be issuing a consultation paper in the coming weeks” and “there is a requirement for a new law because the world of AI is very different” (MediaNama).

  • Still unclear: whether this will be a separate AI law or new rules under the IT Act, 2000, according to MediaNama.

This is a shift. The India AI Governance Guidelines from November 2025 said “a separate law to regulate AI is not needed given the current assessment of risks” and leaned on existing laws plus voluntary measures (India AI Governance Guidelines, PIB).

Which AI rules already apply to Indian businesses?

Two sets of rules already touch businesses that use AI: the 2026 IT Rules amendment on AI-made content, and the Digital Personal Data Protection (DPDP) Rules. The IT Rules amendment has applied to platforms since 20 February 2026; the DPDP Rules were notified on 14 November 2025 and phase in over 18 months. Neither waits for the new consultation paper.

  • IT Rules amendment on AI-made content: in force since 20 Feb 2026. Mainly covers platforms, especially those with over 50 lakh users in India. For you: big platforms must ask you to declare AI-made posts and label them.

  • Draft change to AI labels (April 2026): proposed; comments closed 7 May 2026. Covers platforms. For you: labels on AI visuals may need to stay visible for the whole video.

  • DPDP Rules, Consent Manager rule: applies 12 months after notification. Covers consent managers. Little direct effect on most small firms.

  • DPDP Rules, core duties: apply 18 months after notification (about May 2027). Covers any business handling personal data. For you: clear notices, consent records, security and breach reporting.

  • AI consultation paper: not out yet. Who it covers is not known. For you: read it, then comment.

AI-made content. According to Hogan Lovells, the IT Rules amendment was notified on 10 February 2026 and took effect on 20 February 2026. Permitted AI-made content must be “prominently labelled and embedded with permanent metadata or provenance markers”.

Platforms with more than 50 lakh (5 million) registered users in India “must require users to declare whether content is SGI” (synthetically generated information). Platforms must also act on government or court takedown orders within three hours (Hogan Lovells).

In April 2026, MeitY proposed that labels on AI visuals stay “continuous and clearly visible… throughout the duration of the content” (MediaNama). AZB & Partners’ AI tracker, last updated 8 September 2026, still lists it as a draft (AZB & Partners). This is still unclear: we found no notice making it final.

Personal data. The DPDP Rules were notified on 14 November 2025, with an 18-month phased rollout (PIB). According to Storyboard18, the Consent Manager rule applies after 12 months and the core duties after 18 months (Storyboard18).

Those duties include plain-language notices, a breach report to the Data Protection Board within 72 hours, keeping logs for at least one year and answering complaints within 90 days. India Briefing reports that penalties can go up to ₹250 crore for major violations once full enforcement begins (India Briefing).

Will India’s new AI rules apply to small businesses?

Nobody knows yet, because the draft is not public. On 26 October the minister said the ministry wants input from businesses that use AI, in sectors such as banking, manufacturing and services (MediaNama). On 8 October he said “bulk of the responsibility will have to be taken by the industry” (ThePrint), so in our view small AI users should expect to be part of the conversation.

In our view, three groups should pay the most attention:

  • Businesses with customer-facing AI, such as a WhatsApp chatbot, AI voice agent or AI email replies.

  • Businesses that publish AI-made visuals, such as AI product photos, AI ads or AI voice-overs.

  • Businesses whose automations move customer data between tools, such as a CRM, Google Sheets and an AI model.

If you only use AI privately to draft notes, the risk is lower. This is our interpretation, not a reading of the draft.

What should my business do before the AI draft comes out?

Do the work that pays off whatever the draft says: know which AI tools you use, know what customer data they see, keep a person checking customer-facing AI, label AI-made media and keep a simple record of complaints and fixes. Each step lines up with existing rules or published government principles.

  1. List every AI tool (about 1 hour). Supports accountability in the AI Governance Guidelines.

  2. Map the customer data each tool sees (2 to 3 hours). Supports DPDP notices and security.

  3. Keep a person checking customer-facing AI (ongoing). Supports the minister’s human-control principle.

  4. Label AI-made images, audio and video (10 minutes per post). Supports the 2026 IT Rules amendment.

  5. Keep a complaints and fixes log (30 minutes to set up). Supports the DPDP 90-day complaints rule and the Guidelines on grievance redressal.

Time estimates are our own rough guides for a small team, not official figures. We found no reliable public rupee cost figures for small-business AI compliance, so we have left costs out.

Step 1: List every AI tool you use

Open a Google Sheet with these columns: tool name, what it does, who owns it in the team, which customers it touches, paid or free. CERT-In’s May 2026 blueprint on AI-assisted attacks also recommends keeping AI asset inventories (AZB & Partners).

Ask each team member what they use, including free browser tools. In our experience, teams often find tools nobody signed off on.

Step 2: Map the customer data each tool sees

Add three more columns: what personal data goes in (names, phone numbers, health or payment details), where it is stored and whether customers were told. This is the base for the DPDP notice and security duties due by about May 2027.

Step 3: Keep a person checking customer-facing AI

According to MediaNama, Vaishnaw has spoken about keeping humans in control of AI models, agents and outcomes (MediaNama). In practice, a workflow can look like this:

  • Trigger: a customer sends a WhatsApp message.

  • Action: AI drafts a reply.

  • Condition: if the reply mentions price, refunds, health or legal terms, it waits for staff approval.

  • Outcome: a person sends it, and the decision is logged.

Step 4: Label AI-made images, audio and video

If you post AI-made content on a large platform, it must ask you to declare it. Make labelling a habit now: add “AI-generated” or “Made with AI” to the post and the visual itself.

Step 5: Keep a complaints and fixes log

The AI Governance Guidelines recommend a way for people to report AI-related harms and have them resolved (Guidelines). Under the DPDP Rules, complaints must be answered within 90 days (Storyboard18). A simple form plus a sheet with date, issue, tool involved, fix, closed date is enough to start.

How do I give feedback on India’s AI consultation paper?

When MeitY publishes the paper, it is expected to set a comment window and explain how to send feedback. The minister has said he wants input from citizens, civil society and businesses that use AI (MediaNama). Small businesses can send short, practical comments about cost, clarity and timelines.

  1. Check MeitY’s website and its official social media once a week this month.

  2. Write from your real use. A simple four-line comment works. Who we are: “A 12-person travel company in Chennai.” How we use AI: “An AI chatbot answers booking questions on WhatsApp.” What is unclear or costly: “We don’t know which label to show on AI-made ads.” What we ask for: “A simple label template and a longer deadline for small firms.” (This is an illustration, not a real company.)

  3. Watch the deadline. Comment windows can move; the April 2026 draft’s was extended twice (MediaNama).

What mistakes should businesses avoid while waiting?

The biggest mistake is doing nothing until the final rules arrive. The second is the opposite: buying expensive “AI compliance” tools before anyone knows what the rules require. A simple list, clear labels and a human check cost little and help under any version of the rules.

  • Don’t treat news summaries as the rule. Read the official MeitY text when it is out.

  • Don’t let free AI tools handle customer data without checking where that data goes.

  • Don’t remove AI labels or metadata from content made with AI tools.

  • Don’t forget other markets. If you also sell in the UAE or Singapore, check their rules separately; this post covers India only.

  • Don’t assume “we’re too small”. In our view, the DPDP duties are written for any business that handles personal data, not only large ones.

Want a monthly habit for this? Our Monthly Automation Check-Up is a 15-minute list you can add these checks to.

FAQ

Is there an AI law in India right now?

Not a dedicated AI law yet. The November 2025 AI Governance Guidelines say AI use falls under existing laws, such as IT, data protection and consumer protection laws. In our view, a chatbot that mishandles customer data or misleads buyers can already break current rules, even before any new AI law arrives.

When will India’s AI consultation paper come out?

As of 4 November 2026 we could not find it published; the minister’s last word was “in the coming weeks”. If it lands after you read this, check MeitY’s website for the comment deadline, then re-check your AI tool list against what the paper actually asks for.

Do I need to label AI-generated images on Instagram?

Under the IT Rules amendment in force since 20 February 2026, platforms with more than 50 lakh registered users in India must ask users to declare AI-made content and must label it. So expect large apps to ask you. In our view, adding your own clear “Made with AI” label is a simple habit worth starting now.

Does the DPDP Act apply to my small business chatbot?

If your chatbot collects personal data such as names, phone numbers or booking details, the DPDP duties will apply to you. The core duties take effect 18 months after the rules were notified on 14 November 2025, so about May 2027. A good first step: add a one-line notice to your chatbot’s first message saying what data it collects and why.

What is the fine for breaking India’s data protection rules?

India Briefing reports penalties of up to ₹250 crore for major violations once the Data Protection Board has full enforcement powers in 2027. Actual fines will depend on the type of breach and how it was handled. Until then, know what data you hold, protect it and ask a lawyer about your own situation.

Key takeaways

  • India’s AI consultation paper is expected “in the coming weeks”, but its contents are not public.

  • AI labelling rules for platforms have applied since 20 February 2026.

  • DPDP core duties apply from about May 2027 to any business handling personal data.

  • Five steps work under any outcome: list tools, map data, keep a human check, label AI media, log complaints.

  • Read the official paper and send feedback when it opens.

Where Xvox Labs can help

You don’t need a legal team to get ready. You need a clear view of what you already run. Our Intelligent System Audits go inside your chatbots, automations and AI tools, show what data they touch, flag what is unsafe or broken, and fix it. Humans design the checks. AI does the repetitive part.

If you are working out what AI costs and saves before you change anything, our AI Automation Tools Guide and Affordable AI Automation for Startups are good next reads.

Book a free call and we’ll walk through your AI tools with you.

List it. Label it. Check it. Then let the rules come.

© All right reserved

© All right reserved