bastion

Understanding Data Security

Build this skill

Distinguish personal from sensitive personal data, explain the goals of data protection, and decide what a task needs so that a lost copy exposes as little as possible.

What one medicine order leaves behind

Mrs. Lata Gokhale, 72, opens the Tulsi Health app on the first of the month and reorders her diabetes and blood-pressure tablets. It takes her two minutes. Behind those two minutes, Tulsi Health now holds her name, phone number, home address, a photo of her prescription, the names of her medicines and a reference number for her UPI payment. Data simply means recorded facts like these, whether they sit in a database, a spreadsheet, an email or a paper file. Data security is the work of keeping that data away from people who should not see it, correct and unchanged, and available to the people who need it, for as long as it is kept and no longer.

Tulsi Health app Order #48213 Metformin 500 mg × 60 Amlodipine 5 mg × 30 Prescription photo ✓ Deliver to: Karve Road, Kothrud, Pune Mrs. Lata Gokhale Pay ₹1,240 by UPI Personal data points to one person Name: Lata Gokhale Phone: 98xxx xx417 Address: Karve Road, Kothrud ! Sensitive data serious harm if exposed Medicines: diabetes, BP Prescription photo UPI payment reference One order, six pieces of data Together they say who she is, where she lives, what illness she has and how she pays.

A two-minute order creates six pieces of data. Each piece on its own looks small; together they describe one person's home, health and money.

It is easy to think of data security as a technical subject about servers and passwords. It is really about people. If Mrs. Gokhale's details leak, the harm lands on her: a scam call that knows her medicines, a stranger who knows she lives alone. Tulsi Health holds such details for 2.4 lakh customers, so a single weakness can repeat that harm 2.4 lakh times.

Personal data and sensitive data

Personal data is any information about a person who can be identified from it, either directly or by combining it with other information. A name is personal data, and so is a phone number, a home address, a photo of a face, or a customer ID that Tulsi Health can look up to find the person. Even "the 72-year-old woman on Karve Road who orders on the first of the month" is personal data, because a neighbour could work out who it is.

Some personal data can do far more damage than the rest if it gets out. This is sensitive personal data: health details such as prescriptions and medicines, financial details such as bank accounts, card numbers and payment references, government ID numbers such as Aadhaar and PAN, passwords, fingerprints, and information about children. Think of the difference between your address printed on a wedding invitation and your medical file at the hospital. Many people already know the first; the second can bring shame, discrimination or fraud if a stranger reads it. India's Digital Personal Data Protection Act, 2023 expects organisations to protect the personal data they hold, use it only for the purpose it was given for, and stop keeping it once that purpose is over.

Data that does not point to any person, such as the store's opening hours or the total number of paracetamol strips sold in Nashik last month, is not personal at all. Sort the Tulsi Health items below into the three groups. Ask two questions of each one: could this point to a particular person, and if it did, how badly could they be hurt?

What data protection is trying to achieve

When security people say they are protecting data, they usually mean three goals at once. Confidentiality means only the right people can read the data, so a prescription is seen by the pharmacist preparing the order, not by every employee. Integrity means the data stays correct and nobody changes it without permission, which matters enormously when a changed "5 mg" to "50 mg" could harm a patient. Availability means the data is there when it is needed, such as the prescription history Farida Shaikh needs at 9 pm when a customer asks for a refill.

ConfidentialityOnly the right people read itPharmacist sees the prescriptionIntegrityNobody changes it wrongly5 mg stays 5 mgAvailabilityThere when it is neededHistory opens at 9 pmKeep only what is neededCollect for a clear purposeDelete when the purpose ends

The first three goals protect the data you hold. The fourth reduces how much you hold, which makes the first three easier.

Data protection adds a fourth goal that is easy to forget: hold only what you need. Collecting data for a clear purpose, keeping it only as long as that purpose lasts and knowing where every copy sits is called data minimisation. A shopkeeper who keeps a cash box with only the day's float has less to lose in a robbery than one who keeps a year's takings under the counter. Data works the same way, except that a stolen copy of data never comes back.

How data ends up where it should not

Data exposure means data has become reachable by people who should not have it, whether or not anyone has actually looked yet. A shared-drive folder of supplier contracts accidentally opened to "anyone with the link" is exposed from the moment the setting is changed, even if no stranger has found the link. Exposure is the open door; whether someone walks through is a separate question, and often one you cannot answer, because nobody kept a record of who visited.

Data exposurereachable by thewrong peoplehumanMistakesemail to wrongpersonsettingOpen linksanyone with linkphysicalLost deviceslaptop left ontrainaccessToo much accesseveryone sees allstorageForgotten copiesold exports,backups

Most exposures are not clever attacks. They are ordinary mistakes, settings and leftovers that nobody noticed.

Most exposure starts with ordinary work rather than a skilled attacker. Sameer Joshi emails a salary sheet to "Sameer Patil" instead of a colleague because the email program filled in the wrong name. Someone takes a screenshot of a customer list to ask a question on a staff chat group. A laptop holding an exported spreadsheet is left in an autorickshaw. A test copy of the customer database, made two years ago, still sits on a cloud server that nobody remembers setting up. Each of these is a copy of data living somewhere with weaker protection than the original.

The damage from an exposure depends on two things: how many people's records were exposed, and which details each record contained. A list of 10,000 names and area PIN codes is a nuisance. The same 10,000 rows with prescriptions and payment references attached is a serious harm to every one of those people, and it may require Tulsi Health to inform each of them and report the incident to the authorities.

Keep less, lose less

The surest way to limit an exposure is to decide, before anything goes wrong, how much data a task really needs. Rohan Iyer, the app developer, wants to test a new delivery-route feature on his laptop. The feature needs to know which area each recent order went to, nothing more. The easy option is to export the whole customer table, with every column and every year, because that is one click.

In the lab below you decide what goes into Rohan's export. Then imagine the laptop is lost on the Deccan Queen to Mumbai. The bars show how many people's details of each kind would be in a stranger's hands. Try the mission: keep the route test working while exposing as little as possible.

The lesson carries beyond one spreadsheet. Every copy of data Tulsi Health makes, whether an export, a backup, an email attachment or a screenshot, is something else that has to be protected and eventually deleted. Asking "do we need this, and for how long?" before the copy is made is far cheaper than asking "who has seen this?" after the laptop is gone.

Follow your curiosity.

↑ ↓ to explore · Enter to open · Escape to close