Learn
A2.4.2 | COMMON NETWORK VULNERABILITIES (HL ONLY)
01 | A WEAKNESS, A METHOD AND A CONSEQUENCE
A student sees a convincing message asking them to sign in urgently. Elsewhere, an old server has a known flaw that has never been patched. Both can put a networked system at risk, but the message and the software flaw are different kinds of thing.
A vulnerability is a weakness that can be exploited. A threat is a potential cause of harm. An attack is an attempt to cause harm or gain unauthorised access. An exploit is a method or mechanism that takes advantage of a vulnerability.
The specification’s list includes weaknesses, attack methods and malicious software. We will keep those distinctions clear while covering all ten: DDoS, insecure protocols, malware, MitM, phishing, SQL injection, XSS, unpatched software, weak authentication and zero-day exploits.
This is an HL-only lesson. Your goal is to describe what happens, where the weakness or deception lies and what consequences can follow, rather than memorising a list of frightening names.
02 | WHAT IS THE ATTACK TRYING TO DAMAGE?
Three useful security aims are confidentiality, integrity and availability. Confidentiality concerns who can see information. Integrity concerns protection against unauthorised alteration. Availability concerns whether a service can be used when needed.
If someone reads private assessment records, confidentiality is affected. If someone changes the marks, integrity is affected. If legitimate users cannot open the assessment service, availability is affected. One incident can harm several aims at once.
These aims help you explain impact precisely. “It hacks the computer” tells the reader very little. “It may disclose records the requester is not authorised to read” identifies the outcome and why it matters.
The examples here are conceptual models. Symptoms alone do not prove one attack occurred: a slow page, for example, could also reflect legitimate demand, a technical fault or limited capacity.
03 | DDOS: EXHAUST THE RESOURCES LEGITIMATE USERS NEED
A distributed denial-of-service (DDoS) attack uses traffic or requests from multiple sources to overwhelm a service or its supporting resources. It primarily targets availability, making legitimate access slow or impossible.
The limiting resource might be a network link, connection-state memory, processor capacity or an expensive application operation. An attack need not fill every resource at once to cause disruption. A botnet of compromised devices is one possible source, but not every DDoS uses exactly the same delivery mechanism.
Distribution makes simple source blocking harder because the traffic comes from many sources. A local firewall can help with some filtering, but it cannot by itself restore an upstream connection already saturated before the traffic reaches it.
Heavy legitimate demand can also overload a service. A high request count alone is not proof of malicious intent. The activity below isolates demand and capacity without attempting to identify real traffic or perform any network requests.
The hypothetical service has capacity for 100 equal-work units per second and legitimate demand of 40. These units are illustrative, not a real server benchmark or an allocation guarantee.
04 | INSECURE PROTOCOLS: UNPROTECTED COMMUNICATION
A network protocol defines communication rules, but not every protocol protects confidentiality, integrity or endpoint authenticity. An insecure protocol or configuration may expose sensitive data or allow undetected interference.
For example, an unprotected plaintext login exchange can reveal credentials to someone who can observe that traffic path. Traditional Telnet and ordinary unencrypted FTP are familiar examples of protocols that do not protect their basic credential exchange with transport encryption.
“Plaintext” means the content is not cryptographically protected from a capable observer on that path. It does not mean everyone on the internet automatically sees every packet. Network position and access still matter.
Encrypted and authenticated alternatives reduce relevant risks when correctly configured, but using encryption does not fix every endpoint or application weakness. The specific issue here is insufficient protection of the communication itself.
05 | MALWARE: SOFTWARE WITH A HARMFUL PURPOSE
Malware is software designed to perform harmful or unauthorised actions. It may steal information, monitor activity, damage data, encrypt files for extortion or make a device participate in further attacks.
A virus attaches to a host such as a program or file and spreads when the infected host is executed or otherwise activated. A worm can propagate between systems without needing to attach itself to that kind of host file. A Trojan disguises its harmful role as something legitimate or useful.
Ransomware often encrypts data or disrupts access to demand payment, sometimes alongside data theft. Spyware covertly collects information. These labels can overlap: a malicious package can have several behaviours.
A malicious attachment is not automatically executed simply because a network packet arrived. Execution may involve a user action, an exploited flaw or another mechanism. Likewise, not every type of malware self-replicates, and no single label describes all its delivery routes.
06 | MITM: AN UNTRUSTED PARTY IN THE COMMUNICATION PATH
A man-in-the-middle (MitM) attack, also called an adversary-in-the-middle attack, involves an attacker interposing in communication between parties. The attacker may relay messages while attempting to observe or alter them, with the endpoints believing they are communicating as intended.
Imagine a user reaching what appears to be the correct service through an attacker-controlled intermediary. If the communication lacks effective confidentiality, integrity or endpoint authentication, the intermediary may be able to read credentials or change data.
Being on a network path is not enough to decrypt every properly protected TLS connection. Correct certificate validation and an uncompromised secure configuration resist relevant interception and impersonation. Compromised endpoints, misconfiguration or ignored trust warnings can change the risk.
The essential description includes position and trust: an unauthorised intermediary can interfere with the exchange, rather than simply “a hacker is using the same Wi-Fi”. Normal routers and legitimate proxies are also intermediaries, but their existence alone is not evidence of a MitM attack.
07 | PHISHING: MAKE A PERSON TRUST THE WRONG REQUEST
Phishing uses deceptive communication to persuade someone to reveal information, approve access, send money or run something harmful. A message may impersonate a trusted service, colleague or organisation and create a sense of urgency.
For example, a student receives a fake “school storage full” message that directs them to a false sign-in page. If they enter credentials, the attacker may obtain them. Other phishing schemes ask for a permission grant, a one-time code or approval rather than just a password.
The communication can arrive by email, text, messaging, voice or another channel. Spear phishing is targeted using details relevant to a person or group. Good spelling and a familiar-looking logo do not establish authenticity.
A phishing page can itself use HTTPS: encryption to the wrong endpoint does not make that endpoint honest. The weakness being exploited may involve misplaced trust or a deceptive workflow rather than a bug in the network protocol.
08 | SQL INJECTION: INPUT CHANGES A DATABASE INSTRUCTION
A web application might accept a search term and query a database. SQL injection occurs when untrusted input is incorporated into SQL in a way that lets it change the intended database instruction rather than remain ordinary data.
The problem is the boundary between data and instructions. If software constructs a query unsafely from input, an attacker may influence what the database executes. Depending on the flaw and database permissions, the result can include unauthorised reading, changes, deletion or authentication bypass.
This is not the same as the user merely typing SQL words into a textbox. Whether input can alter execution depends on how the application builds and executes the query. Ordinary validation alone is not a reliable substitute for keeping query structure separate from data.
Parameterised queries are a central defensive approach: the intended SQL structure and supplied values are handled separately. We develop countermeasures next; here you should recognise that the relevant execution target is the database through the vulnerable application. No attack payload is needed to explain that mechanism.
09 | XSS: UNTRUSTED CONTENT BECOMES BROWSER CODE
Cross-site scripting (XSS) occurs when an application lets attacker-controlled content execute in a user’s browser as part of the trusted site’s context. The browser may then perform actions or expose information available to that context.
For example, a discussion page displays an untrusted contribution in an unsafe way. Another visitor opens the page and the browser treats some of that content as executable code rather than harmless text. This can affect the victim even though the browser arrived at the legitimate site.
Stored XSS persists in content later served to visitors. Reflected XSS returns attacker-controlled request content in a vulnerable response. DOM-based XSS arises from unsafe client-side processing of untrusted data. These descriptions can overlap in real systems.
Consequences can include altered displayed content, unauthorised in-session actions or access to exposed information. It is not correct to promise that every XSS can read every cookie: cookie attributes, browser rules and application design constrain particular effects. Context-appropriate output handling and safe rendering are important defences.
10 | SQL INJECTION AND XSS: FOLLOW THE DIFFERENT TARGETS
Both involve an application mishandling untrusted data. The important difference is where data becomes instructions and what executes them. SQL injection affects database instructions; XSS affects browser execution in a site context.
Use the selector to follow the conceptual sequence. All displayed content is inert explanatory text. The activity does not execute a query, inject a script or connect to another system.
11 | UNPATCHED SOFTWARE: A KNOWN FIX HAS NOT BEEN APPLIED
A patch is a software update that addresses a problem, which can include a security flaw. Unpatched software can remain exposed to known vulnerabilities when the applicable security update has not been installed.
An example is a service running a version with a publicly documented flaw and an available vendor fix. An attacker may try to exploit that weakness if the vulnerable function is reachable and other required conditions are met. Old version numbers alone do not prove every particular exploit will work.
Patch management involves inventory, risk assessment, testing, deployment and verification. Unsupported software can create additional problems because suitable fixes may no longer be available. These operational decisions matter even if the network firewall is functioning.
Do not confuse “unpatched” with “zero-day”. They can both involve flaws, but an already available fix that has not been applied is different from the no-fix or previously unknown situation associated with zero-day exploitation.
12 | WEAK AUTHENTICATION: INADEQUATE CONFIDENCE IN IDENTITY
Authentication checks an identity claim. Weak authentication makes it too easy to gain access as someone else, for example through default credentials, easily guessed passwords, reused compromised passwords or a poorly protected account-recovery process.
An attacker may obtain a credential through guessing, a previous breach, phishing or malware. The success conditions vary; the existence of a login page does not mean its identity checks are strong. Reusing one password can connect the risk across multiple services.
Authentication is different from authorisation, which decides what an identified user may do. A properly authenticated student reading staff records because of faulty access checks is an authorisation failure, not necessarily weak authentication.
Multi-factor authentication can improve identity assurance, but its effectiveness depends on the method and workflow, including recovery. It is not a promise that users can never be deceived or that sessions and endpoints cannot be compromised.
13 | ZERO-DAY EXPLOITS: LITTLE OR NO PREPARATION TIME
A zero-day vulnerability is commonly used to describe a newly discovered or previously unknown flaw for which defenders have had little or no opportunity to prepare, often with no vendor fix available when it is exploited. A zero-day exploit is a mechanism taking advantage of that flaw.
For example, attackers use a previously unknown flaw in a network service before an effective vendor patch is available. Defenders may initially lack a specific signature or fix even though other controls could still reduce exposure or detect unusual behaviour.
Terminology varies with disclosure and patch status. The key comparison is with a known flaw whose fix is already available but has not been applied. Once a vulnerability is understood and a patch is released, failing to install that patch becomes a patch-management issue.
“Zero-day” does not mean the attack lasts exactly one day, is impossible to detect or defeats every security measure. Reduced reachability, least privilege, monitoring and containment can still limit some consequences.
14 | COMBINE THE IDEAS WITHOUT MAKING THEM IDENTICAL
A real incident can involve several of these concepts. A deceptive message might persuade a user to run a Trojan, which steals account information that is then misused. Each term describes a different part of the chain.
Another incident might begin with an unpatched application flaw and lead to data theft. It does not need phishing at all. A DDoS can deny service without stealing a password or executing code on the target.
Choose a conceptual incident below and step through its stages. These are possible sequences, not a rule that every attack follows the same path.
15 | COMPARE MECHANISMS AND LIKELY CONSEQUENCES
| Item | Defining mechanism or weakness | Possible consequence |
|---|---|---|
| DDoS | Multiple sources exhaust service or network resources | Loss of availability |
| Insecure protocols | Insufficient confidentiality, integrity or endpoint protection | Observed credentials or altered communication where the attacker has access |
| Malware | Harmful or unauthorised software behaviour | Theft, damage, disruption or further compromise |
| MitM | Unauthorised intermediary interferes with an exchange | Disclosure or alteration where protections fail |
| Phishing | Deception induces an unsafe action or disclosure | Credential/access theft, malware delivery or fraud |
| SQL injection | Input changes intended database instructions | Unauthorised database access or changes |
| XSS | Untrusted content executes in browser site context | Altered page behaviour or unauthorised user-context actions |
| Unpatched software | Relevant known fixes are absent | Continued exposure to applicable known flaws |
| Weak authentication | Identity checks or credentials are inadequate | Unauthorised use of another identity |
| Zero-day exploit | Exploits a newly unknown/unfixed flaw with little preparation opportunity | Compromise whose specific fix may initially be unavailable |
These are possible consequences, not a prediction that every instance causes every outcome. Effective permissions, protocol protections and other conditions change the impact.
16 | IDENTIFY THE DEFINING FEATURE
Each situation highlights one concept. Real incidents can contain several, so select the term matching the specific feature described rather than guessing from the presence of a website or a login box.
17 | COMMON CONFUSIONS
| Claim | Correction |
|---|---|
| A slow website proves DDoS | Legitimate load and faults can also cause delay; distribution and hostile resource exhaustion define the attack |
| HTTPS means the page is honest | It can protect a connection to a deceptive endpoint |
| SQL injection and XSS execute in the same place | SQL injection targets database instruction handling; XSS targets browser site-context execution |
| Every malware type replicates itself | Replication and other behaviours vary |
| An authenticated user can access everything | Authorisation must still restrict permitted actions |
| Zero-day means unpreventable for exactly 24 hours | It concerns discovery/preparation/fix circumstances, not a fixed duration or universal success |
Strong descriptions state the mechanism and conditions. Avoid claims that encryption never helps, that one defence guarantees protection, or that a symptom establishes the complete cause.
18 | DESCRIBE THE PROCESS AND ITS IMPACT
For describe, give the main characteristics in clear subject-specific language. A useful pattern is: define the concept, explain the relevant weakness or action, then state a plausible consequence in the scenario.
For example: “A DDoS uses requests or traffic from many sources to exhaust service resources. Legitimate requests may then be delayed or rejected, reducing availability.” Or: “SQL injection occurs when untrusted input can alter a database instruction through a vulnerable application, potentially exposing or changing records.”
For a distinction: “Phishing deceives a person into an unsafe action, whereas MitM places an unauthorised party in the communication path. A phishing incident may later deliver malware, but these terms describe different mechanisms.”
Use because, resulting in and so that where they clarify cause and effect. The next lesson develops the countermeasures; here, be precise about what goes wrong and where.