Basic Concept: Context window DoS attacks flood an LLM ' s context with obfuscated content to exhaust processing resources or manipulate model behavior. Attackers may hide large amounts of text behind Unicode characters like emojis. CompTIA SecAI+ Study Guide identifies prompt filtering as the primary defense against input-based attacks on LLMs.
Why D is Correct: A prompt filter inspects incoming inputs before they reach the LLM, detecting and blocking malicious content including obfuscated text hidden behind Unicode characters or emojis. By analyzing input structure, character counts, hidden content, and encoding anomalies, prompt filters can identify and reject attacks that attempt to abuse the context window, preventing resource exhaustion.
Why A is Wrong: Fraud detection systems are designed to identify fraudulent transactions or activities in structured data contexts. They are not designed to inspect LLM prompt structures for obfuscated content attacks on context windows.
Why B is Wrong: LLM-as-a-judge uses a secondary LLM to evaluate the quality or safety of another model ' s outputs. It operates post-generation and cannot prevent a DoS attack that occurs during input processing before output is generated.
Why C is Wrong: Pattern recognition can identify known attack patterns but requires the attack to match pre- learned patterns. Novel obfuscation techniques using Unicode or emoji hiding may evade pattern-based detection without dedicated prompt filtering logic.

Explanation:
See Explanation

Basic Concept: This is a Performance-Based Question (PBQ) - a simulation requiring interactive protocol and cipher selection in the actual exam. It tests the candidate ' s ability to select appropriate encryption protocols for different AI system components and apply proper data security techniques for sensitive AI data.
Key Concept - Encryption by Component: For data in transit between AI components, TLS 1.3 is the current standard protocol. For API communications, HTTPS with TLS 1.3 and certificate pinning is appropriate. For database connections used by AI systems, TLS with strong cipher suites such as AES-256-GCM should be applied. For data at rest in model stores and training data repositories, AES-256 encryption is standard. For authentication tokens and keys, RSA-2048 or ECC P-256 are appropriate asymmetric options.
Key Concept - Data Techniques: For sensitive AI training data that has been modified, tokenization replaces sensitive values with non-sensitive tokens. Masking obscures sensitive fields in outputs. Hashing provides one-way verification for data integrity.
Reference: CompTIA SecAI+ Study Guide Domain 2 covers encryption requirements for AI system data security. Candidates should know standard encryption protocols (TLS 1.3, AES-256), when to apply each cipher type, and which data security techniques such as masking, tokenization, and encryption at rest are appropriate for different AI system data categories and sensitivity levels.

Basic Concept: AI systems interact with different resource layers including models, data stores, and infrastructure. Controlling what data an AI system can access requires implementing access controls at the data layer. CompTIA SecAI+ Study Guide differentiates between model access, data access, and network access controls for AI systems.
Why D is Correct: Data access controls govern what data resources an AI system can interact with, including which databases, tables, and records it can read or modify. To control an AI system ' s ability to list database tables, the administrator must implement data access controls that define precisely which tables the AI can enumerate and query, following the principle of least privilege for data interactions.
Why A is Wrong: Agentic AI access refers to permissions granted to autonomous AI agents to perform actions and use tools. It is a broader concept about what an AI agent can do operationally rather than a specific data-layer access control mechanism.
Why B is Wrong: A Network Access Control List controls network traffic at the IP and port level, determining which hosts can communicate with which network resources. It operates at the network layer and cannot enforce fine-grained control over which database tables an AI system is allowed to list.
Why C is Wrong: Model access controls govern who and what can interact with the AI model itself - who can query it, update it, or access its parameters. This is distinct from data access, which controls what the model can read from data stores during operation.

Basic Concept: AI systems used in employment contexts such as job screening carry significant regulatory risk. For a multinational company operating in or serving markets covered by the EU AI Act, compliance with this binding regulation is mandatory to avoid substantial fines. CompTIA SecAI+ Exam Objectives cover AI regulatory compliance under Domain 4.
Why B is Correct: The EU AI Act explicitly classifies AI systems used for employment screening, candidate evaluation, and worker management as high-risk AI applications. These systems are subject to strict compliance requirements including mandatory conformity assessments, human oversight, transparency obligations, and registration. Non-compliance can result in fines up to 30 million euros or 6% of global annual turnover. A multinational company implementing AI job screening must reference the EU AI Act as the primary compliance obligation.
Why A is Wrong: ISO AI standards such as ISO 42001 are voluntary management system standards. While useful for best practices, they do not carry legal enforcement power and adherence does not prevent regulatory fines from binding legislation like the EU AI Act.
Why C is Wrong: Corporate policy is an internal governance document that sets organizational standards. It cannot supersede external legal obligations and following only corporate policy does not protect against fines from regulatory bodies enforcing the EU AI Act.
Why D is Wrong: NIST AI RMF is a voluntary American risk management framework. While excellent for AI risk governance, it is not a binding regulation and does not address the legal compliance requirements that generate fines from regulatory authorities in jurisdictions covered by the EU AI Act.

Basic Concept: Prompt injection occurs when malicious content embedded in user input manipulates an LLM
' s behavior, causing it to leak sensitive data, bypass restrictions, or execute unintended actions. Preventing such attacks requires mechanisms that inspect and filter content at the prompt level. CompTIA SecAI+ covers LLM-specific security controls extensively.
Why A is Correct: Guardrails are purpose-built controls that inspect, filter, and constrain both input prompts and output responses in LLM systems. They can detect sensitive data patterns such as credit card numbers, block prompt injection payloads, enforce content policies, and prevent the model from processing or outputting restricted information. Guardrails are the primary LLM-native defense against prompt injection as cited in the CompTIA SecAI+ Study Guide.
Why B is Wrong: Antivirus software detects known malware signatures in files and executables. It does not inspect or understand the semantic content of LLM prompts and cannot detect or block prompt injection attacks.
Why C is Wrong: A WAF operates at the HTTP layer inspecting web requests and responses against rule sets.
While it can block some patterns, it lacks the contextual intelligence to understand LLM prompt semantics and cannot prevent sophisticated injection attacks.
Why D is Wrong: Role-based access control manages who can access which resources. It controls authorization but does not inspect the content of prompts to prevent injection attacks once a user has legitimate access.