Learn
5.3.2 | KEEPING DATA SAFE
01 | SECURITY USES SEVERAL LAYERS
No single measure protects against every threat. Layered protection combines controls with different roles.
For example, authentication checks a user’s identity, access levels limit their actions, anti-malware looks for malicious software and a secure connection protects transmitted content.
When explaining a solution, link its mechanism to its benefit: say what it does and why that reduces a particular risk.
02 | ACCESS LEVELS
Access levels determine what an account is permitted to read, change or administer. Users should receive the rights needed for their role rather than unrestricted access.
| School role | Illustrative access |
|---|---|
| Student | Read lesson resources and edit their own work |
| Teacher | Manage resources and assessment information for authorised classes |
| Administrator | Perform approved system-management tasks |
Benefit: limiting privileges reduces what an unauthorised person or mistaken user can change through that account.
Access levels do not authenticate identity. They control permitted actions after identity and account context have been established.
03 | USERNAME AND PASSWORD
Authentication checks a claimed identity. A username identifies an account; a password supplies secret information used to authenticate it.
Where passwords are used, long, unique passwords reduce the risk from guessing and reuse of a password exposed elsewhere. The username itself is not normally the secret.
A password can still be disclosed through phishing or captured on a compromised device. Knowing a password is therefore not proof that the person entering it is the intended user.
The syllabus includes username/password authentication. Real services may also offer alternatives such as passkeys; the important principle is checking identity through an appropriate mechanism.
04 | BIOMETRICS
Biometric authentication uses a measured human characteristic, such as a fingerprint or facial features, to help verify a user.
The system compares a captured sample with enrolled biometric information using an appropriate matching process.
Benefit: the user does not need to remember that characteristic as a password. Limit: matching can wrongly reject or accept a user, and biometric information is sensitive.
Biometrics are not guaranteed to be impossible to imitate. Some systems also use a fallback credential or additional checks.
05 | TWO-STEP VERIFICATION
Two-step verification requires an additional verification step beyond the initial login. A familiar example is entering a password and then using an authenticator-generated code or approving a legitimate sign-in request.
Benefit: a stolen password alone may be insufficient to access the account.
Two-factor authentication uses evidence from different factor categories, such as something known and something possessed. Simply entering a username and password is not two-factor authentication.
An attacker can still try to trick a user into revealing a code or approving a request. Do not share verification codes or approve a sign-in you did not initiate.
06 | ANTI-MALWARE
Anti-malware detects, blocks or deals with malicious software. The syllabus includes anti-virus and anti-spyware.
It can compare files with known signatures and use other detection methods, such as identifying suspicious behaviour. Depending on the tool, detected content may be blocked, quarantined or removed.
Benefit: it reduces the likelihood that harmful software runs or remains active. Updating the tool and its detection information helps it recognise more threats.
It cannot guarantee detection of every new or disguised threat. A successful scan is not proof that all files are safe.
07 | AUTOMATING SOFTWARE UPDATES
Security updates can fix known weaknesses in an operating system or application. Automatic updates reduce reliance on a user remembering to obtain and apply those fixes.
Cause and consequence: installing a relevant fix removes or reduces an exploitable weakness, limiting opportunities for attackers to use it.
Some updates need a restart or other completion step. Automation must still be checked to ensure updates were successfully applied.
Updates reduce risk; they do not guarantee that every weakness is known or repaired.
08 | CHECK SPELLING, TONE AND CONTEXT
Deceptive communications may contain unusual spelling, inconsistent branding or a tone that pressures the user to act urgently.
Examples: a supposed support message demands a password, a payment instruction unexpectedly changes, or a request warns of immediate punishment unless a link is followed.
These clues can prompt a pause and further verification. Check an unexpected important request through a known, independent channel rather than relying only on the supplied contact details.
Good spelling does not prove legitimacy. Attackers can write convincing messages, and genuine messages can contain mistakes.
09 | CHECK THE URL ATTACHED TO A LINK
A link’s displayed text can differ from its actual destination. Inspect the destination where possible and identify the host name carefully.
| Teaching address | Host that would be contacted |
|---|---|
| https://school.example.org/login | school.example.org |
| https://school.example.org.attacker.invalid/login | school.example.org.attacker.invalid |
| https://attacker.invalid/school.example.org/login | attacker.invalid |
These are illustrative domains, not links to follow. On different devices, inspecting a link can use different controls. A known bookmark or independently entered trusted address can avoid following a suspicious message link.
HTTPS protects the connection to the named host; a deceptive site can also use HTTPS. A reassuring scheme alone does not establish the organisation’s identity.
10 | FIREWALLS
A firewall monitors and filters network traffic according to rules. Rules can allow or block communication based on information such as addresses, ports, protocols or connection state.
Benefit: it can restrict unwanted connections and reduce exposure of services. A host firewall protects an individual device; a network firewall can control traffic across a network boundary.
A firewall is not the same as anti-malware. Allowed traffic can still contain a deceptive message or harmful content, and configuration affects what it blocks.
11 | PRIVACY SETTINGS
Privacy settings control aspects of what information is shared, collected or visible to others. Examples include audience restrictions and application permissions.
Benefit: reducing unnecessary exposure can limit information available for profiling, impersonation or social engineering.
Settings do not guarantee that recipients cannot copy information. They also do not necessarily erase copies already obtained or stop all tracking mechanisms.
12 | PROXY SERVERS
A proxy server acts as an intermediary between a client and another service. In a school, a configured forward proxy can handle requests from students’ devices.
Send a request through the configured proxy
↓ Apply policy
Allow, block or forward according to its capabilities
↓ Forward an allowed request
Return a response through the intermediary
A proxy can enforce access policy, log requests or cache suitable resources. It can also mean that the destination sees the proxy’s outward network address instead of the client’s direct address.
Limits: a proxy does not automatically encrypt traffic or make the user anonymous to the proxy operator. HTTPS content cannot simply be inspected by every proxy without suitable additional arrangements.
13 | SSL AND MODERN TLS
The specification names Secure Sockets Layer (SSL). SSL is the historical secure-connection protocol; modern web connections use its successor, Transport Layer Security (TLS), although the label “SSL” is still commonly used.
The key syllabus concept is protection of data exchanged over a network through a secure connection.
During setup, certificate checks help authenticate the server for the requested host and the parties establish cryptographic protection. Exchanged content is encrypted and its integrity is protected.
Benefit: an observer cannot simply read sensitive details from the network exchange, and tampering can be detected. This does not prevent a dishonest site from receiving data a user chooses to submit.
14 | MATCH PROTECTION TO THREAT
| Problem | Relevant protection | Why it helps |
|---|---|---|
| Unauthorised account use | Authentication and two-step verification | Require appropriate evidence of identity |
| Excessive rights | Access levels | Limit permitted actions |
| Malicious software | Anti-malware and updates | Detect threats and repair known weaknesses |
| Deceptive request | Check message context and URL; verify independently | Avoid trusting a forged communication |
| Unwanted connections | Firewall rules | Restrict network access |
| Sensitive data in transit | Secure SSL/TLS concept used by HTTPS | Encrypt and protect the exchange |
| Unnecessary public exposure | Privacy settings | Reduce shared information |
| School browsing policy | Configured proxy | Control permitted requests |
Several methods can address one scenario. A phishing message may be reduced by awareness, site filtering and strong authentication, but none should be described as an absolute guarantee.
15 | BACKUP: A USEFUL ADDITION
Separate, recoverable backups can support recovery after deletion, failure or ransomware. Recovery should be checked by testing that needed files can be restored.
A backup does not itself prevent a phishing attack or undo stolen information. Synchronisation that propagates deletion is not automatically an independent recoverable backup.
This extends the listed protections by considering recovery as well as prevention.
16 | COMMON MISTAKES
- Authentication and access levels are identical: checking identity differs from granting rights.
- A username is a second authentication factor: it identifies the account.
- Anti-malware detects everything: new or disguised threats may be missed.
- Perfect spelling proves a message genuine: convincing phishing exists.
- HTTPS proves a site honest: it protects the connection.
- A firewall and proxy have identical roles: filtering rules differ from intermediary requests.
- A proxy automatically encrypts everything: protection depends on the communication setup.
- Privacy settings prevent every disclosure: copied or previously shared data can remain.
OPTIONAL READING
See the NCSC two-step verification guidance and explanation of SSL and its TLS successor for further context.