Skip to content
Home » Breach & Incident Response » The Small Business Owner’s Guide to Incident Response in India

The Small Business Owner’s Guide to Incident Response in India

Illustration showing a cyberattack on a laptop with warning alerts and a firefighter responding, representing incident response and recovery for small business cybersecurity.

The call usually comes at a bad time. A staff member cannot open the Tally company, a client asks about an email your office never sent, or a ransom note sits on the server’s desktop. What you do in the next few hours decides how much you lose. An incident response plan for a small business is simply that decision made in advance: who does what, in which order, and whom you must tell. This guide is written for Indian firms, especially CA practices with one Windows server and a part-time IT vendor, and it follows Indian law rather than the US advice most guides repeat.

Not legal advice. This post summarises the CERT-In Directions, the IT Act and the DPDP Act as they stood in October 2026. Confirm how they apply to your firm with a legal advisor.

Three clocks that start when you notice

Most incident response advice online assumes American law: state notification rules, the FBI, credit monitoring. None of it applies here. In India, three separate duties can start running the moment you notice an incident:

WhoWhenWhat
CERT-InWithin 6 hours of noticing, or of being told. In force since 2022.Report specified cyber incidents, including unauthorised access, ransomware and data breaches. Email [email protected] or call 1800-11-4949. Send what you know; details can follow.
Data Protection Board and each affected personFrom mid-May 2027: without delay, plus a detailed report to the Board within 72 hours.Any personal data breach, under the DPDP Act 2023 and DPDP Rules 2025. Until then, IT Act section 43A applies.
Police (cyber crime)At once if money has moved.Call the 1930 helpline and file a complaint at cybercrime.gov.in. Call your bank too.

Sources: the CERT-In Directions of 28 April 2022 (section 70B(6) of the IT Act), the DPDP Rules 2025 as published by PIB and the National Cyber Crime Reporting Portal. CERT-In’s FAQ on the Directions treats a firm as a body corporate and says the reporting duty cannot be handed to someone else, so your IT vendor’s report does not cover you. Ignoring the Directions can mean a fine of up to ₹1 crore, up to a year’s imprisonment, or both (section 70B(7), as amended in November 2023).

What counts as an incident

Failed password attempts and connections your firewall drops happen to every server on the internet. They are noise, though you should keep the logs. These are not noise:

  • A successful sign-in from an address or at an hour that nobody can explain.
  • A user account nobody created, or someone added to the Administrators group.
  • Files that suddenly cannot be opened, renamed extensions, or a ransom note.
  • Clients receiving emails or payment requests “from you” that you did not send.
  • The Windows Firewall switched off, the security log cleared, or antivirus disabled without anyone on your side doing it.

Deciding where noise ends and an incident begins is a judgement call. When in doubt, take advice quickly and lean towards reporting.

Step 1: Notice it, and write down the time

The six hours start when you notice, not when you are ready, so the first action is to note the date and time. Then tell the one person who is in charge of incidents. In a small practice that is usually a partner; name a deputy as well, for the day the partner is travelling.

Bear in mind that the visible moment is rarely the start. Mandiant’s M-Trends 2026 put the global median time between an intrusion and its discovery at 14 days in 2025. Somebody may have been signing in to your server for a while before the ransom note appeared. The signs to look for, with the Windows event IDs behind them, are in 7 signs your Windows server may have been compromised and how to tell if someone is trying to break in.

Step 2: Contain it, and report to CERT-In

Containment means stopping the damage from spreading. It does not mean fixing everything.

  1. Isolate, do not power off. Unplug the network cable or turn off Wi-Fi on the affected server or PC. Shutting it down wipes evidence still held in memory; CISA’s ransomware guide advises powering down only if you cannot disconnect it from the network.
  2. Cut the way in. If Remote Desktop is open to the internet, remove the port-forward on the office router. Disable any account you suspect, and change its password from a different, clean computer.
  3. Do not delete, wipe or “clean up” yet. Do not pay or negotiate a ransom before you have taken expert and legal advice; government agencies advise against paying.
  4. Report to CERT-In within six hours. Email [email protected] (toll-free phone 1800-11-4949, fax 1800-11-6969) with what you know: firm name and contact, when you noticed, what happened, which systems, what you have done so far. Say clearly what is still being investigated. Keep a copy and note the time you sent it.
  5. If money has moved, call 1930 now and file at cybercrime.gov.in. Speed matters for getting a transfer held. Call your bank’s fraud line as well.
  6. From mid-May 2027, if personal data was involved, the DPDP Rules also require telling the Data Protection Board and each affected person without delay. This runs alongside the CERT-In report, not instead of it.

Our free CERT-In playbook, The Six-Hour Rule, sets out an hour-by-hour plan for this step and includes a report template you can fill in today. It is on the Security Resources page, in PDF and Word.

Step 3: Find out what happened

Now the difficult question: how far did it go? Which accounts were used, which systems were touched, and was any client data copied? This is usually your IT vendor’s job, and it can take days. Before anything is changed, save the Windows logs to a fresh USB drive (not your backup drive), from an elevated PowerShell window on the affected server:

New-Item -ItemType Directory -Path E:\IR -Force
wevtutil epl Security E:\IR\Security.evtx
wevtutil epl System E:\IR\System.evtx
wevtutil epl Application E:\IR\Application.evtx

Replace E: with your USB drive’s letter, and unplug it as soon as the copy finishes. Useful entries to look for: 4624 with logon type 10 (a Remote Desktop sign-in), 4720 (an account created), 4732 (someone added to a local group such as Administrators), 7045 in the System log (a new service installed) and 1102 (the security log cleared). Our post on event ID 4625 shows how to read the failed sign-ins that often come first.

CERT-In expects you to keep logs of all ICT systems for a rolling 180 days in India and hand them over with a report. A firm that has been keeping them can answer “when did it start?” in an hour. A firm that has not may never know.

Step 4: Remove the attacker and close the door

Remove what the attacker left behind: unknown accounts, services and scheduled tasks, and remote-support tools nobody installed on purpose. Then close the way they came in. If that was Remote Desktop with a guessed password, the fix is to take RDP off the internet, not to change the port. Prixl Guard lets invited staff reach the server through an encrypted tunnel so RDP can be closed; a VPN or RD Gateway does the same job. How to restrict RDP access safely walks through the options.

Reset every password the attacker could have seen, not only the obvious one: Windows accounts, Tally users, email, the income-tax and GST portal logins stored on that machine. After ransomware, rebuilding the server from clean media is usually safer than trying to clean it.

Step 5: Recover

Restore from a backup the attacker could not reach: a drive that was disconnected, or cloud storage with immutability turned on. Check that the backup pre-dates the intrusion, not just the ransom note. Bring back the Tally company data and client folders first, check them, and only then reconnect the server to the network.

If ransomware is involved, you will face a choice between restoring, rebuilding from scratch and the attacker’s offer. Paying does not guarantee you get your data back, and it does not remove the attacker’s access. A tested, offline backup is what lets you skip this conversation. How ransomware gets into Windows servers covers backups that survive an attack.

Step 6: Tell the people who need to know, then learn from it

  • CERT-In: follow up as the facts become clear, and answer its questions within the time it sets.
  • Data Protection Board (from mid-May 2027): the detailed report within 72 hours of becoming aware of the breach.
  • Clients: a CA owes clients confidentiality under the Chartered Accountants Act, and clients’ PAN, bank and GST details are exactly what attackers want. Decide with your legal advisor what to tell them and when; from May 2027 the DPDP Act will require telling each affected person.
  • Your insurer, if you have cyber cover. Check the policy now for its notice deadline and whom it wants you to call first.

Within two weeks, sit down with the partners and the IT vendor. How did they get in? Were there warning signs nobody saw? What slowed you down? Write the answers in a register and change the plan. The cost side of this is covered in the real cost of a data breach for a small business, and the legal exposure when a hijacked server is used against others in your server is sending spam.

Mistakes that make it worse

  • Switching the server off, or formatting it, before the logs are saved.
  • Plugging the backup drive into the infected server “to check it”.
  • Changing passwords from the compromised machine, where a keylogger may be watching.
  • Waiting for the full picture before reporting to CERT-In. Report what you know.
  • Coordinating over the firm’s email account, if that account may be compromised. Use phones.

Your plan on one page

You do not need a long document. Fill in these headings, print two copies, and keep one at a partner’s home. If the server is encrypted, the copy on the server is no use.

  • Who is in charge: the partner who leads, a deputy, and their mobile numbers.
  • Contacts, written down offline: IT vendor (and their out-of-hours number), CERT-In ([email protected], 1800-11-4949), cyber crime helpline 1930, your bank’s fraud line, legal advisor, insurer.
  • How you will notice: which alerts reach a phone, including at night and on Sundays.
  • First actions per scenario: ransom note, stolen password, fraudulent email to clients.
  • Backups: where they are, who can restore them, and the date of the last restore test.
  • Logs: where the Windows logs are kept, and for how long (CERT-In: 180 days; one year under the DPDP Rules from May 2027).
  • Whom to tell, and by when: the three clocks above, plus clients and the insurer.

Then rehearse it for half an hour with the partners: “the server is encrypted at 9 pm on a Saturday in July. Who does what?” The gaps show up quickly. For the method behind these steps, NIST’s SP 800-61 Revision 3 (2025) is the standard reference.

Where Prixl Warden fits

Every step above depends on Step 1, and Step 1 depends on someone seeing the signs. Prixl Warden watches the Windows event logs on your server and emails you when something serious happens: a brute-force campaign, fast or slow, a sign-in outside working hours, a new account or a change to the Administrators group, a new service, the audit log being cleared, or the firewall being switched off. A daily digest at 18:00 IST and a Monday summary cover the rest, and event history is kept for 30 days. It will not respond for you, but it shortens the time between “it started” and “we noticed”.

Start your 30-day free trial. No card is needed.

Related reading

Leave a Reply

Your email address will not be published. Required fields are marked *