Zero Trust and Secure Account Management
Decide whether to allow, challenge or block a request using Zero Trust principles, and manage an account through joining, moving, review and deactivation.
Being inside the office is not a password
For years, Tulsi Health's systems worked on one simple assumption: if a computer is connected to the head office network in Kothrud, it must belong to staff, so it can reach the shared drive without further questions. Then a visiting salesman plugged his laptop into a spare network socket in the meeting room, and his laptop could see every invoice folder. Nothing was hacked; the network simply trusted anyone inside it. Zero Trust is the security approach that drops this assumption. Its motto is "never trust, always verify": no person, device or connection is trusted just because of where it is, and every request to reach data is checked on its own merits. The other half of the story is managing accounts well over their whole life, because Zero Trust is only as good as the accounts and permissions it checks.
The old castle and the airport
The older approach is often called the castle-and-moat model. A strong wall (the firewall, a device that filters traffic at the edge of a network) keeps outsiders out, and everything inside the wall is trusted. Its weakness is that once anyone gets inside, whether by a phished password, a plugged-in laptop or malware on one computer, they can wander freely. Today the "inside" is also hard to define: staff check email on phones at home, the app and its database run in the cloud, and Rathi Softech connects remotely from Indore to update MedLedger.
An airport works differently. Your ticket and ID are checked at the entry gate, your bag is scanned at security, your boarding pass is scanned again at the gate, and the crew checks your seat number at the aircraft door. Being inside the terminal never gets you onto a plane by itself. Zero Trust treats every system the same way.
Zero Trust repeats the check at every point where something valuable can be reached, not only at the front door.
Three principles of Zero Trust
The first principle is verify explicitly. Every request is judged using everything known about it: who the user is (checked with a password plus a second factor such as an authenticator app), which device they are using and whether it is updated and belongs to the company, where they are connecting from, and what they are asking for. The second principle is least privilege: even a verified user gets only the access their job needs, and often only for as long as they need it. The third is assume breach: design as if an attacker is already inside somewhere, so that one compromised laptop or account cannot reach everything, and so that activity is logged and watched.
Zero Trust does not remove the wall. It adds a check at every door inside it, so getting past the wall is no longer enough.
The second half of the first principle is continuous verification. Instead of checking once at the morning login and trusting the person for the rest of the day, the system keeps watching for changes in risk. If Sameer's session suddenly comes from a new country, a new device, or starts downloading thousands of customer records, the system can ask him to verify again, limit what he can do, or end the session. Most of the time nobody notices these checks. They step in only when something about the request no longer looks like the person it claims to be.
Try being the gatekeeper for a few real requests. For each one, decide what a Zero Trust system should do.
The account lifecycle: joiners, movers and leavers
Zero Trust checks every request against the permissions an account holds, so those permissions must be right. The account lifecycle is the life of an account from creation to removal, usually described in three events. A joiner is a new person who needs an account with the right role on day one, and no more. A mover changes job inside the organisation, and must gain the new role's access and lose the old one, otherwise privilege creep sets in. A leaver departs, and every account and access they had must be switched off promptly.
Each stage is tied to an HR event. Access that is not tied to an event, such as a 'temporary' admin right, is the access that gets forgotten.
The safest lifecycle is driven by events the organisation already records. When HR adds Riya as an intern, her account is created with the Intern role. When her internship ends on 31 December, the same HR record triggers removal. Linking accounts to these events means nobody has to remember to do it.
Permission reviews and account deactivation
Even with good processes, access drifts. A permission review (also called an access review) is a regular check, perhaps every three months, in which each manager sees the list of what their team members can reach and confirms or removes each item. Arjun would see that a cashier still belongs to the Accounts group from a year-end two years ago, and remove it. Reviews also catch accounts that belong to nobody: test accounts, old vendor logins and shared logins with no named owner.
Account deactivation is switching off an account so it can no longer be used. For a leaver it should happen by the end of their last working day, or immediately if they leave on bad terms. Deactivation is usually done before deletion: the account is disabled and its active sessions are ended, but its records and files are kept for a while because the business and the law may need them. It also covers more than the main login. Shared passwords the person knew must be changed, company phones and laptops collected, and access to cloud tools, vendor portals and group chats removed. A society secretary leaving office hands over every key, not just the main gate key, and the locks for anything they kept a copy of are changed.
Put the steps of Riya's account lifecycle in the right order, from her first day to the week after she leaves.