Skip to main content

Threats and Attacks

Understanding cybersecurity threats and attacks is essential for CPAs working in IT audit, advisory, and assurance. Organizations face a constantly evolving threat landscape — from disgruntled insiders to sophisticated nation-state actors — and each threat requires a different set of controls. As a CPA, you must be able to classify threat agents, identify attack types and techniques, explain how cyber-attacks progress through stages, and evaluate the cybersecurity risks specific to cloud, IoT, and mobile environments. This section covers types of threat agents and their motivations, categories of attacks, specific cyber-attack techniques, the stages of a cyber-attack (Cyber Kill Chain), cloud cybersecurity risks, IoT cybersecurity risks, mobile technology risks, threat modeling and threat landscape, and determining threats to third-party connections and organizational systems.

info

The ISC exam tests threats and attacks primarily at the Remembering and Understanding level, with select topics tested at the Application and Analysis levels. You should be able to classify threat agents, identify attack types and techniques, explain the stages of a cyber-attack, and determine specific cybersecurity threats for a given scenario.


Types of Threat Agents

A threat agent (also called a threat actor) is any individual, group, or entity that has the potential to cause harm to an organization's information systems. Threat agents are classified by their relationship to the organization and by their motivations.

Internal vs. External Threats

ClassificationDescriptionExamples
Internal threatsIndividuals with authorized access to the organization's systems who misuse that accessEmployees, contractors, temporary workers, business partners with VPN access
External threatsIndividuals or groups without authorized access who attempt to breach the organization's defensesHackers, organized crime groups, nation-state actors, hacktivists

Internal threats are particularly dangerous because insiders already have legitimate credentials, knowledge of internal systems, and physical access to facilities — making detection more difficult.

Categories of Threat Actors

Threat ActorDescriptionMotivationSophistication
Script kiddiesInexperienced attackers who use pre-built tools and scriptsCuriosity, bragging rights, minor disruptionLow
HacktivistsIndividuals or groups who attack systems to promote a political or social causeIdeological, political activismLow to moderate
Organized crimeCriminal groups that conduct cyber-attacks for financial profitFinancial gain (fraud, extortion, data theft)Moderate to high
Nation-state actorsGovernment-sponsored groups that conduct cyber espionage or attacksEspionage, sabotage, geopolitical advantageVery high
Advanced Persistent Threats (APT)Highly skilled groups (often nation-state sponsored) that maintain prolonged, stealthy access to a targetLong-term intelligence gathering, IP theftVery high
Insider threatsEmployees or contractors who intentionally or accidentally cause harmFinancial gain, revenge, negligenceVaries
CompetitorsRival organizations seeking proprietary informationCompetitive advantage, trade secret theftModerate
Exam Tip

When a question asks you to classify a threat agent, focus on two factors: (1) whether the actor is internal or external to the organization, and (2) whether the actor is nation-state sponsored or non-nation-state sponsored. A disgruntled employee is an internal, non-nation-state threat. A government cyber-warfare unit is an external, nation-state threat.

Example: Bear Co. discovers that a former contractor retained VPN credentials after their engagement ended and used those credentials to access sensitive financial data. This is an internal threat (the contractor had authorized access) even though the person is no longer actively employed.


Types of Attacks

Cyber-attacks are categorized by the method used to compromise systems, data, or personnel.

Physical Attacks

Physical attacks exploit the physical environment rather than digital vulnerabilities:

AttackDescription
Tailgating (piggybacking)Following an authorized person through a secured door without presenting credentials
Dumpster divingSearching through discarded materials (trash, recycling) for sensitive information
Shoulder surfingObserving someone's screen or keyboard to capture passwords or sensitive data
Theft of devicesStealing laptops, USB drives, or mobile devices to access stored data
VandalismPhysically damaging equipment, cables, or infrastructure

Distributed Denial of Service (DDoS)

A Distributed Denial of Service (DDoS) attack overwhelms a target system, server, or network with a flood of traffic from multiple sources, making it unavailable to legitimate users. DDoS attacks do not typically steal data — they disrupt availability. Key characteristics:

  • Traffic originates from many compromised devices (a botnet)
  • Targets the availability principle of the CIA triad
  • Can cost organizations significant revenue during downtime
  • Often used as a distraction while other attacks occur simultaneously

Malware

Malware (malicious software) is any software designed to damage, disrupt, or gain unauthorized access to computer systems:

Malware TypeDescriptionPropagation
VirusMalicious code that attaches to a legitimate program and executes when the host program runsRequires user action (opening a file, running a program)
WormSelf-replicating malware that spreads across networks without user interactionAutomatic — exploits network vulnerabilities
Trojan horseMalware disguised as legitimate software that performs hidden malicious functionsUser downloads and installs what appears to be legitimate software
RansomwareEncrypts the victim's data and demands payment for the decryption keyTypically via phishing emails or exploited vulnerabilities
SpywareSecretly monitors user activity and collects sensitive informationBundled with downloads or installed via exploits
RootkitHides deep within the operating system to provide persistent, undetected accessInstalled after initial compromise; very difficult to detect
KeyloggerRecords keystrokes to capture passwords, credit card numbers, and other sensitive dataInstalled via trojan or physical access

Social Engineering

Social engineering attacks manipulate human psychology rather than exploiting technical vulnerabilities:

AttackDescriptionMedium
PhishingMass emails impersonating a trusted entity to trick recipients into revealing credentials or clicking malicious linksEmail
Spear phishingTargeted phishing directed at a specific individual or organization using personalized informationEmail
WhalingSpear phishing targeting senior executives (CEO, CFO)Email
VishingVoice phishing — using phone calls to trick victims into revealing sensitive informationPhone
SmishingSMS phishing — using text messages to deliver malicious linksText message
PretextingCreating a fabricated scenario to gain a victim's trust and extract informationAny medium
BaitingLeaving infected media (USB drives) in public areas hoping someone will plug them inPhysical
warning

Social engineering is the most common initial attack vector in data breaches. Technical controls (firewalls, encryption) cannot fully protect against social engineering because these attacks target people, not systems. Organizations must implement security awareness training as a critical control.

Web Application Attacks

Web application attacks target vulnerabilities in web-based software:

  • SQL injection — inserting malicious SQL code into input fields to manipulate databases
  • Cross-site scripting (XSS) — injecting malicious scripts into web pages viewed by other users
  • Cross-site request forgery (CSRF) — tricking authenticated users into performing unintended actions
  • Session hijacking — stealing or predicting session tokens to impersonate authenticated users

Mobile Device Attacks

Mobile-specific attacks include:

  • Exploiting vulnerabilities in mobile apps
  • Man-in-the-middle attacks on unsecured Wi-Fi networks
  • Malicious apps distributed through unofficial app stores
  • SMS-based attacks (smishing, SIM swapping)
  • Exploiting Bluetooth or NFC vulnerabilities

Cyber-Attack Techniques

Beyond the general categories of attacks, the CPA exam requires knowledge of specific techniques used to carry out cyber-attacks:

TechniqueDescriptionImpact
Buffer overflowSending more data to a memory buffer than it can hold, causing the program to overwrite adjacent memory and potentially execute malicious codeCode execution, system crash
Mobile codeCode (e.g., JavaScript, ActiveX, Java applets) that is downloaded and executed on a user's system, potentially without the user's knowledgeUnauthorized actions on the user's device
Cross-site scripting (XSS)Injecting client-side scripts into web pages that are then executed in other users' browsersSession hijacking, credential theft, defacement
SQL injectionInserting malicious SQL statements into application queries to read, modify, or delete database contentData breach, data manipulation, authentication bypass
Race conditionExploiting the timing gap between a security check and the use of the checked resourcePrivilege escalation, unauthorized transactions
Covert channelAn unauthorized communication path that transfers information in violation of security policyData exfiltration without detection
Replay attackCapturing and retransmitting valid authentication data (e.g., login credentials, session tokens) to gain unauthorized accessAuthentication bypass
Return-oriented programming (ROP)Chaining together short sequences of existing code (called "gadgets") to execute malicious instructions without injecting new codeBypasses defenses that prevent code injection
Example: An attacker discovers that MAS Inc.'s customer portal does not properly validate user input. The attacker enters ' OR 1=1 -- into the login field, which modifies the SQL query to return all records from the user table. This is a SQL injection attack that bypasses authentication.
caution

Buffer overflow and SQL injection are among the most historically common and dangerous attack techniques. Both exploit a failure to properly validate input. Input validation is a critical application-level control that prevents many of these techniques.


Stages of a Cyber-Attack (Cyber Kill Chain)

A cyber-attack is not a single event — it unfolds through a sequence of stages. Understanding these stages helps organizations detect and stop attacks at each phase. Two commonly referenced frameworks describe these stages:

Lockheed Martin Cyber Kill Chain

StageDescriptionExample
ReconnaissanceAttacker researches the target — gathering information about systems, employees, and vulnerabilitiesScanning Gies Co.'s public-facing servers for open ports
WeaponizationAttacker creates a deliverable payload (e.g., malware embedded in a document)Building a malicious PDF exploiting a known vulnerability
DeliveryAttacker transmits the weapon to the targetSending a spear-phishing email to Gies Co.'s CFO
ExploitationThe payload triggers and exploits a vulnerabilityThe malicious PDF executes code when opened
InstallationMalware is installed on the target system to establish persistenceA backdoor is installed on the CFO's workstation
Command & Control (C2)The compromised system establishes a communication channel back to the attackerThe backdoor connects to an external server for instructions
Actions on ObjectivesThe attacker achieves their goal (data exfiltration, destruction, manipulation)Financial data is exfiltrated to an external server

Alternative Stage Model (Exam-Referenced)

The CPA exam also references these stages:

StageDescription
ReconnaissanceGathering information about the target (network scanning, social media research, OSINT)
Gaining accessExploiting a vulnerability or using stolen credentials to enter the system
Escalation of privilegesMoving from a low-privilege account to administrator or root access
Maintaining accessInstalling backdoors, rootkits, or other persistence mechanisms to retain access
Network exploitationMoving laterally across the network to access additional systems and data
Covering tracksDeleting logs, modifying timestamps, and removing evidence of the intrusion
Exam Tip

Both stage models describe the same concept — a cyber-attack progresses through sequential phases. A key takeaway is that early detection is critical. If an organization detects an attacker during the reconnaissance phase, it can prevent all subsequent stages. The further an attacker progresses, the more damage they can inflict and the harder it is to contain.


Cloud Cybersecurity Risks

Cloud computing introduces unique cybersecurity risks that do not exist (or are less prominent) in traditional on-premises environments:

RiskDescription
Shared responsibility gapsMisunderstanding of which controls are the CSP's responsibility vs. the customer's, leading to unprotected areas
Misconfigured storageCloud storage buckets (e.g., AWS S3) left publicly accessible due to incorrect permission settings
API vulnerabilitiesInsecure application programming interfaces used to manage cloud services can be exploited
Data co-minglingIn multi-tenant environments, one customer's data may be logically stored alongside another's, creating isolation risks
Insider threats at CSPEmployees of the cloud service provider may have access to customer data
Account hijackingStolen cloud credentials can give attackers full control over the organization's cloud environment
Data residency and sovereigntyData stored in foreign jurisdictions may be subject to different legal frameworks
Vendor lock-inDifficulty migrating away from a CSP can reduce an organization's ability to respond to security concerns
Example: Illini Security migrates its financial reporting application to a public cloud but fails to restrict access to a cloud storage bucket containing customer records. The storage is left with default public-read permissions, exposing sensitive data to anyone on the internet. This is a misconfigured storage risk.

IoT Cybersecurity Risks

The Internet of Things (IoT) refers to physical devices (sensors, cameras, thermostats, industrial equipment) that connect to the internet and exchange data. IoT devices dramatically expand an organization's attack surface.

RiskDescription
Default credentialsMany IoT devices ship with factory-default usernames and passwords that are never changed
Lack of patchingIoT manufacturers may not provide firmware updates, leaving known vulnerabilities unpatched
Limited processing powerMany IoT devices lack the computing resources to run encryption or security software
Large attack surfaceOrganizations may deploy thousands of IoT devices, each representing a potential entry point
Botnet recruitmentCompromised IoT devices are often recruited into botnets used to launch DDoS attacks
Insecure communicationIoT devices may transmit data without encryption, exposing it to interception
Physical accessibilityIoT devices are often deployed in public or uncontrolled locations where they can be physically tampered with
note

The Mirai botnet attack (2016) demonstrated how compromised IoT devices — primarily security cameras and routers with default passwords — could be used to launch massive DDoS attacks. This illustrates why default credential management is a critical IoT control.


Mobile Technology Risks

Mobile devices (smartphones, tablets) present unique cybersecurity challenges because they operate outside the organization's physical security perimeter:

RiskDescription
BYOD (Bring Your Own Device)Personal devices used for work may lack enterprise security controls
Jailbreaking / rootingRemoving manufacturer security restrictions bypasses built-in protections and exposes the device to malware
Unsecured Wi-FiConnecting to public Wi-Fi networks exposes data to man-in-the-middle attacks
App vulnerabilitiesMalicious or poorly coded apps can compromise data stored on the device
Device loss or theftLost or stolen devices may provide physical access to corporate data and applications
Outdated operating systemsUsers who do not update their devices miss critical security patches
Data leakageCorporate data can be inadvertently shared through personal cloud storage, messaging apps, or screenshots
Example: An employee at Kingfisher Industries uses their personal smartphone (BYOD) to access the corporate email system. The employee connects to a public Wi-Fi hotspot at a coffee shop without using a VPN. An attacker on the same network performs a man-in-the-middle attack and intercepts the employee's email credentials.

Threat Modeling and Threat Landscape

Threat Modeling

Threat modeling is a structured process for identifying, quantifying, and addressing security threats to an application or system. It answers four key questions:

  1. What are we building (or protecting)?
  2. What can go wrong?
  3. What are we going to do about it?
  4. Did we do a good enough job? Common threat modeling methodologies:
    MethodologyDescriptionFocus
    STRIDECategorizes threats into six types: Spoofing, Tampering, Repudiation, Information disclosure, Denial of service, Elevation of privilegeApplication-level threats
    PASTAProcess for Attack Simulation and Threat Analysis — a seven-stage risk-centric methodologyBusiness risk alignment
    Attack treesGraphical representations of possible attacks, with the goal at the root and attack methods as branchesSpecific attack scenarios
    DREADRates threats by Damage, Reproducibility, Exploitability, Affected users, DiscoverabilityRisk prioritization

Threat Landscape

The threat landscape refers to the complete set of potential and active threats facing an organization at any given time. It is shaped by:

  • The organization's industry (healthcare, finance, government)
  • The organization's geographic location and applicable regulations
  • The organization's technology stack and attack surface
  • Current geopolitical events and trending attack methods
  • The organization's public profile and value as a target Organizations should regularly assess their threat landscape to ensure controls remain aligned with actual risks.

Determining Threats to Third-Party Connections

Organizations do not operate in isolation. Connections with customers, vendors, and partner organizations create additional cybersecurity threats:

Connection TypeThreatsControls
Vendor connectionsCompromised vendor systems can serve as an entry point (supply chain attack); vendors with excessive access may introduce riskVendor risk assessments, SOC reports, least-privilege access, network segmentation
Customer portalsExternal-facing systems are targets for web application attacks, credential stuffing, and DDoSWeb application firewalls (WAF), rate limiting, multi-factor authentication
Partner integrationsAPI-based integrations can expose data if APIs are insecure or partner systems are compromisedAPI security testing, mutual TLS authentication, data minimization
Supply chainMalicious code or backdoors inserted into third-party software or hardwareSoftware composition analysis, hardware integrity verification, vendor due diligence
Example: Illini Entertainment integrates its ticketing system with a third-party payment processor via an API. If the payment processor is compromised, attackers could potentially pivot through the API connection to access Illini Entertainment's customer database. This is a supply chain risk that requires vendor risk assessment and network segmentation controls.

Determining Threats to On-Premise and Cloud Systems

CPAs must be able to analyze the specific cybersecurity threats to an organization's applications, networks, and connected devices:

Asset CategoryKey ThreatsAnalysis Considerations
On-premise applicationsUnpatched vulnerabilities, insider threats, physical access attacksPatch management processes, access controls, physical security
Cloud-based applicationsMisconfiguration, identity attacks, API exploitationCloud security posture management, IAM policies, CSP SOC reports
NetworksMan-in-the-middle attacks, unauthorized access, DDoS, lateral movementNetwork segmentation, firewall rules, intrusion detection/prevention
Mobile devicesDevice loss, malware, unsecured connections, data leakageMDM solutions, encryption, remote wipe, app whitelisting
IoT devicesDefault credentials, firmware vulnerabilities, botnet recruitmentNetwork isolation, credential management, firmware updates
Exam Tip

When analyzing threats in a scenario, consider the specific context of the organization. A healthcare organization faces different primary threats (patient data theft, ransomware disrupting care) than a financial services firm (wire fraud, trading system manipulation). The threat analysis must be tailored to the organization's industry, data types, and technology environment.


Summary

TopicKey Takeaway
Threat agentsClassified as internal/external and by motivation (financial, political, espionage); nation-state actors are the most sophisticated
Types of attacksInclude physical attacks, DDoS, malware, social engineering, web application attacks, and mobile attacks
Cyber-attack techniquesBuffer overflow, SQL injection, XSS, replay attacks, and others exploit specific technical vulnerabilities — most are prevented by input validation
Stages of a cyber-attackAttacks progress through reconnaissance → access → privilege escalation → persistence → exploitation → covering tracks
Cloud risksShared responsibility gaps, misconfigured storage, API vulnerabilities, and data co-mingling are primary concerns
IoT risksDefault credentials, inability to patch, and large attack surfaces make IoT devices attractive targets for botnets
Mobile risksBYOD, unsecured Wi-Fi, device loss, and jailbreaking create entry points outside the organization's perimeter
Threat modelingStructured methods (STRIDE, PASTA) to identify and prioritize threats; the threat landscape is the full picture of active and potential threats
Third-party threatsVendor connections, supply chain attacks, and partner integrations require vendor risk assessments and network segmentation

Practice Questions

  1. Gies Co. is a mid-sized manufacturing company. A security analyst discovers that an attacker has been present in the network for six months, slowly exfiltrating intellectual property related to a proprietary manufacturing process. The attacker appears to be well-funded and uses custom malware not seen in any public threat database. How should this threat agent be classified?
  2. MAS Inc. receives a report that its cloud-based HR application has a storage bucket containing employee Social Security numbers that is accessible to anyone with the URL — no authentication required. What type of cloud cybersecurity risk does this represent, and what is the most likely root cause?
  3. Kingfisher Industries operates a smart building with 2,000 IoT sensors monitoring temperature, lighting, and security cameras. The IT security team has never changed the factory-default credentials on these devices. During a recent audit, several sensors were found to be sending traffic to unknown external IP addresses. What IoT-specific risk has materialized, and what stage of the cyber-attack lifecycle does the observed behavior represent?
    Answers
  4. This is an Advanced Persistent Threat (APT), likely nation-state sponsored. The key indicators are: (a) prolonged, stealthy presence (six months), (b) sophisticated custom malware, (c) targeting of intellectual property, and (d) substantial resources and patience. APTs are external, nation-state or state-sponsored threat actors focused on long-term intelligence gathering.
  5. This represents a misconfigured storage risk — the most common cloud security failure. The root cause is a shared responsibility gap: the CSP provides the tools to secure storage buckets, but it is the customer's (MAS Inc.'s) responsibility to configure proper access controls. Someone either left the default public-access setting unchanged or incorrectly configured the bucket's permissions.
  6. The materialized risk is botnet recruitment due to default credentials. The devices were compromised because factory-default passwords were never changed — a fundamental IoT security failure. The observed behavior (traffic to unknown external IPs) indicates the devices have reached the Command & Control (C2) stage of the cyber-attack lifecycle, where compromised devices communicate with the attacker's infrastructure to receive instructions. :::