Certified - AI Security Audio Course
The AI Security & Threats Audio Course is a comprehensive, audio-first learning series focused on the risks, defenses, and governance models that define secure artificial intelligence operations today. Designed for cybersecurity professionals, AI practitioners, and certification candidates, this course translates complex technical and policy concepts into clear, practical lessons. Each episode explores a critical aspect of AI security—from prompt injection and model theft to data poisoning, adversarial attacks, and secure machine learning operations (MLOps). It also covers global standards and regulatory guidance, including the NIST AI Risk Management Framework, ISO/IEC 23894, and emerging organizational policies around transparency, accountability, and continuous monitoring.
Episodes
-
Episode 1 — Course Overview & How to Use This Prepcast 14.09.2025 21mThis opening episode provides a structured orientation to the AI Security and Threats Audio course series, helping listeners understand what the program covers and how to best engage with the material. The overview defines the scope of AI security by placing it within the broader context of cybersecurity and risk management, while clarifying the distinctive elements that make AI-specific security necessary. It explains how the episodes are organized, moving from foundational principles through attack surfaces, defenses, governance frameworks, and advanced considerations. The episode also outli... -
Episode 2 — The AI Security Landscape 14.09.2025 23mThis episode defines the AI security landscape by mapping the assets, attack surfaces, and emerging threats that distinguish AI from classical application security. It introduces critical components such as training data, model weights, prompts, and external tools, explaining why each must be protected as an asset. The relevance for certification exams lies in understanding how these components shift trust boundaries and create new risks compared to traditional software systems. The episode emphasizes that adversaries target AI differently, often exploiting natural language, data poisoning, or... -
Episode 3 — System Architecture & Trust Boundaries 14.09.2025 21mThis episode explains the architecture of AI systems, breaking down their stages and components to show how trust boundaries shift across the lifecycle. Training, inference, retrieval-augmented generation (RAG), and agent frameworks are introduced as discrete but interconnected environments, each with distinct risks. For exam relevance, learners are expected to identify these architectural elements, describe where threats occur, and understand how adversaries exploit them. The discussion highlights how traditional security boundaries—such as network segmentation or user authentication—must be... -
Episode 4 — Data Lifecycle Security 14.09.2025 23mThis episode examines data lifecycle security, covering the journey of data from collection and labeling through storage, retention, deletion, and provenance management. It explains why data is the foundation of AI system reliability and how its misuse or compromise undermines security objectives. For certification preparation, learners are introduced to key definitions of provenance, integrity, and retention policies, while understanding how regulatory requirements drive data governance practices. The episode situates data lifecycle security as both a technical and compliance necessity, bridg... -
Episode 5 — Prompt Security I: Injection & Jailbreaks 14.09.2025 22mThis episode introduces prompt injection and jailbreaks as fundamental AI-specific security risks. It defines prompt injection as malicious manipulation of model inputs to alter behavior and describes jailbreaks as methods for bypassing built-in safeguards. For certification purposes, learners must understand these concepts as new categories of vulnerabilities unique to AI, distinct from but conceptually parallel to classical injection attacks. The discussion highlights why prompt injection is considered one of the highest risks in generative AI systems, as it can expose sensitive data, trigge... -
Episode 6 — Prompt Security II: Indirect & Cross-Domain Injections 14.09.2025 22mThis episode examines indirect and cross-domain prompt injections, which expand the attack surface by embedding malicious instructions in external sources such as documents, websites, or email content. Unlike direct injection, where the attacker provides inputs to the model directly, these threats exploit retrieval or integration features that feed information into the AI system automatically. Learners preparing for certification exams must understand the mechanics of these attacks, which occur when contextual data bypasses normal user input validation and reaches the model unchecked. The rele... -
Episode 7 — Content Safety vs. Security 14.09.2025 20mThis episode explains the distinction and overlap between content safety and security in AI systems, a concept often emphasized in both professional practice and certification exams. Content safety refers to filtering or moderating outputs to prevent harmful or offensive material, while security focuses on protecting systems and assets from adversarial manipulation or data loss. Although they are related, treating them as identical can cause organizations to miss critical risks. Learners must grasp why an AI model can pass content safety tests yet still be vulnerable to prompt injection, data... -
Episode 8 — Data Poisoning Attacks 14.09.2025 24mThis episode introduces data poisoning as a high-priority threat in AI security, where adversaries deliberately insert malicious samples into training or fine-tuning datasets. For exam readiness, learners must understand how poisoning undermines model accuracy, introduces backdoors, or biases outputs toward attacker goals. The relevance of poisoning lies in its persistence, as compromised models may behave unpredictably long after training is complete. Definitions such as targeted versus indiscriminate poisoning, as well as the concept of trigger-based backdoors, are emphasized to ensure candi... -
Episode 9 — Training-Time Integrity 14.09.2025 21mThis episode covers training-time integrity, focusing on the assurance that data, processes, and infrastructure used in model development remain uncompromised. Learners preparing for exams must understand that threats at this stage include data tampering, corrupted labels, or manipulated hyperparameters. Unlike inference-time attacks, which target deployed models, training-time compromises affect the foundation of the model itself, potentially embedding vulnerabilities that persist throughout the lifecycle. The exam relevance lies in being able to identify how training-time risks manifest and... -
Episode 10 — Privacy Attacks 14.09.2025 27mThis episode introduces privacy attacks in AI systems, focusing on techniques that reveal sensitive or personal information from training data or model behavior. Learners must be able to define key attack types, such as membership inference—determining whether a specific record was included in training—and model inversion, where attackers reconstruct approximate training inputs. The exam relevance lies in understanding not only the mechanics of these attacks but also their implications for regulatory compliance and user trust. Privacy risks are especially significant in domains such as healthc... -
Episode 11 — Privacy-Preserving Techniques 14.09.2025 26mThis episode explores privacy-preserving techniques designed to reduce the risk of sensitive information exposure in AI systems while maintaining utility of the models. Learners must understand concepts such as anonymization, pseudonymization, and data minimization, which limit identifiable information in training sets. Differential privacy is introduced as a mathematical framework that injects statistical noise into data or queries, providing measurable privacy guarantees. Federated learning is also explained as a decentralized training method that keeps raw data on user devices, mitigating r... -
Episode 12 — Model Theft & Extraction 14.09.2025 29mThis episode addresses model theft and extraction, highlighting how adversaries can replicate or steal valuable AI models. Model theft occurs when proprietary weights or architectures are exfiltrated, while model extraction involves querying an exposed API repeatedly to reconstruct decision boundaries or functionality. For exam purposes, learners must be able to distinguish between these two concepts and describe the potential impacts, which include intellectual property loss, competitive disadvantage, and undermining of security guarantees. These risks make model theft an enterprise-level con... -
Episode 13 — Adversarial Evasion 14.09.2025 29mThis episode introduces adversarial evasion, a class of attacks in which maliciously crafted inputs cause AI systems to misclassify or behave incorrectly. For exam purposes, learners must be able to define adversarial examples, explain why they are often imperceptible to humans, and distinguish them from poisoning attacks, which occur during training. Evasion attacks take place at inference time and undermine confidence in model reliability. The episode covers historical research origins in image recognition and extends to natural language and audio domains, illustrating the cross-modal nature... -
Episode 14 — RAG Security I: Retrieval & Index Hardening 14.09.2025 29mThis episode explores retrieval-augmented generation (RAG) security, focusing on retrieval and index hardening as foundational defenses. RAG combines language models with external document retrieval, which improves factual grounding but introduces risks. Learners preparing for exams must understand how poisoning of indexes, adversarial queries, and tampered retrieval sources can compromise model outputs. The episode explains why vector databases, document indexes, and retrievers are critical assets requiring protection, emphasizing that compromised retrieval pipelines can lead to misinformatio... -
Episode 15 — RAG Security II: Context Filtering & Grounding 14.09.2025 21mThis episode continues exploration of RAG security by examining context filtering and grounding as defenses for reliable outputs. Learners must understand context filtering as the screening of retrieved documents before they are passed to a model, ensuring that malicious or irrelevant content is excluded. Grounding is defined as aligning model outputs to trusted sources, improving accuracy and reducing hallucination. For exam purposes, mastery of these definitions and their application to AI security is critical, as context and grounding directly affect confidentiality, integrity, and trustwor... -
Episode 16 — Agents as an Attack Surface 14.09.2025 26mThis episode introduces AI agents as a new and growing attack surface, highlighting how their autonomy and tool integration create unique risks. Agents differ from single-response models by persisting through plan-and-act loops, chaining multiple steps, and invoking external tools or APIs. For certification purposes, learners must understand that these design features expand the system boundary, exposing new trust assumptions and vulnerabilities. Risks include prompt injection, privilege escalation, excessive resource consumption, and data exfiltration when agents interact with connected servi... -
Episode 17 — Secrets & Credential Hygiene 14.09.2025 22mThis episode addresses secrets and credential hygiene, emphasizing their critical role in preventing leaks and privilege misuse in AI systems. Secrets include API keys, tokens, passwords, and configuration values embedded in prompts or environments. Learners preparing for exams must understand that secrets frequently appear in AI workflows, often stored insecurely or accidentally revealed in logs or outputs. Credential hygiene practices ensure that secrets are generated securely, stored in vault systems, rotated regularly, and protected against unauthorized access. The exam relevance lies in i... -
Episode 18 — AuthN/Z for LLM Apps 14.09.2025 25mThis episode explores authentication (AuthN) and authorization (AuthZ) for large language model (LLM) applications, highlighting their importance in managing identities and permissions. Authentication verifies that a user or system is who they claim to be, while authorization defines what actions or resources they are allowed to access. For certification readiness, learners must understand the difference between these two concepts, recognize their application in AI contexts, and describe how least privilege is enforced across sessions and scopes. The exam relevance lies in knowing how access c... -
Episode 19 — Output Validation & Policy Enforcement 14.09.2025 30mThis episode examines output validation and policy enforcement as mechanisms for controlling what AI systems produce before results are delivered to users or downstream processes. Output validation ensures that responses conform to expected formats or structures, such as JSON schemas, while policy enforcement applies organizational rules that block disallowed or unsafe outputs. For exam purposes, learners must understand how these layers complement input validation, creating a defense-in-depth strategy that limits both harmful behavior and misuse. Definitions of allow lists, deny lists, and st... -
Episode 20 — Red Teaming Strategy for GenAI 14.09.2025 29mThis episode introduces red teaming as a structured method for probing generative AI systems for vulnerabilities, emphasizing its importance for both exam preparation and real-world resilience. Red teaming involves adopting an adversarial mindset to simulate attacks such as prompt injection, data leakage, or abuse of system integrations. For learners, understanding red team goals, rules of engagement, and reporting requirements is essential to certification-level mastery. The relevance lies in recognizing how red teaming complements audits and testing pipelines by uncovering weaknesses that or...
Popular in
The podcast also appears in the podcast charts of these countries.