AI is no longer waiting at the edge of the business. It is already inside workflows, products, security operations and everyday decisions. The rapid integration of AI across organizations creates enormous opportunity, but it also raises a practical question for the cybersecurity profession: Are our organizational capabilities, technical skills, and security standards keeping pace with the systems we rely on today?

The release of the ISC2 AI Security Certification Pilot Exam Outline helps answer that question. It offers a first look at the body of knowledge ISC2 is validating for a new professional credential focused on securing AI systems, managing AI-related threats and mitigating AI risk. It also signals a broader shift: AI security is becoming a defined professional practice spanning governance, data protection, secure architecture, operations, incident response and security testing.

How the ISC2 AI Security Certification Pilot Exam Was Developed

The forthcoming pilot exam is the next step in ISC2’s standards-based certification development process, informed by input from a pool of nearly 10,000 volunteer professionals worldwide through Job Task Analysis surveys, exam item development and expert review. That practitioner-led approach gives the outline practical weight because AI security is being shaped by professionals managing risk in fast-changing environments.

AI security certification is intended for someone responsible for securing the design, deployment and maintenance of AI and machine learning systems throughout the AI lifecycle. The role includes managing cybersecurity risks associated with AI-enabled services and extending beyond technical defense to governance, ethics, risk-based guidance and enterprise collaboration.

What Topics Are Covered on the ISC2 AI Security Certification Pilot Exam?

The pilot exam outline organizes the knowledge, skills and abilities expected of AI security professionals across six domains:

  • Artificial Intelligence Concepts and Security Fundamentals
  • AI Governance, Ethics, Risk and Compliance
  • AI Data Management, Privacy and Model Security
  • Secure AI Architecture, Engineering and Identity and Access Management (IAM)
  • Secure AI Operations and Incident Response
  • AI Threats, Testing and Lifecycle Management

Together, these domains show how quickly the work of securing AI has expanded. AI security is not limited to defending a model from attack. It requires fluency in AI concepts, machine learning, generative AI architectures, prompt engineering, retrieval-augmented generation, autonomous agents and model performance risks such as hallucinations, drift and model collapse.

The outline also moves into governance and risk, reinforcing that responsible AI security requires more than technical controls. Professionals are expected to understand major AI regulations, standards and frameworks; support AI policies and use-case inventories; assess risk; and enable human oversight, audit practices and accountability. That breadth reflects the reality organizations are already navigating: AI risk requires coordination among cybersecurity, data science, legal, compliance, procurement, business owners and executive leadership.

AI Data, Model Security and Privacy Topics Covered by the Certification

One of the most notable aspects of the outline is how clearly it elevates data and model security as core responsibilities. AI systems depend on data quality, provenance, privacy and integrity, and the outline includes the governance, protection and privacy controls needed to manage those risks across the AI lifecycle.

It also reinforces that models themselves are valuable and vulnerable assets. Professionals are expected to understand how to protect model confidentiality, integrity and availability as AI introduces new attack surfaces across training data, model weights, prompts, retrieval systems, deployment pipelines and autonomous workflows.

Secure AI Architecture, Operations and Incident Response Exam Topics

The ISC2 AI Security Certification Pilot Exam Outline also makes clear that securing AI requires architectural discipline from the start. Topics include Zero Trust, defense-in-depth, privacy by design, threat modeling, secure integration, deployment pipeline security, validation checkpoints, rollback procedures and guardrails for autonomous AI behavior.

Operational readiness is where that architecture meets reality. Cybersecurity professionals must be prepared to monitor AI activity, detect runtime threats and misuse, maintain visibility into AI assets and respond to incidents such as prompt injection, data poisoning, model theft and agent misuse. The outline also recognizes the need for AI-specific containment, forensics, recovery and post-incident review.

Why the ISC2 AI Security Certification Pilot Exam Matters

The pilot exam phase is where this practitioner-led work becomes more tangible. Approximately 3,000 qualified volunteers are expected to sit for the pilot exam from mid-December through late January, helping validate the exam against the capabilities identified through the development process. If successful, the certification is expected to become operational and launch in 2027.

For cybersecurity professionals, the outline offers a timely roadmap for where AI security practice is heading. For employers, it provides insight into the capabilities teams may need to govern, build, deploy and defend AI responsibly. And for the profession, it represents meaningful progress toward a common, vendor-neutral standard for AI security expertise.

AI is changing what organizations can do, and cybersecurity practice must keep up with what now needs to be secured. The ISC2 AI Security Certification Pilot Exam Outline highlights a powerful point: securing AI requires professionals who can connect technical safeguards, governance expectations, operational resilience and ethical responsibility across the AI lifecycle. With this pilot exam outline, ISC2 is helping define that pragmatic path for the AI era.