ISC2 has begun developing a new AI Security certification, marking its latest effort to address the growing demand for cybersecurity professionals with expertise in securing artificial intelligence systems.
The organization is opening the development process to volunteers worldwide, inviting both ISC2 members and non-members to participate in defining the certification's body of knowledge, validating exam objectives, and helping develop exam questions.
According to ISC2 CISO Jon France, the certification is intended to provide a vendor-neutral credential focused on the security challenges surrounding AI technologies rather than specific products or platforms. The effort follows increasing enterprise adoption of generative AI, AI-powered security tools, and autonomous AI agents, all of which have introduced new attack surfaces and governance concerns.
The certification development will follow ISC2's established process for creating new credentials. The first phase will identify the knowledge, skills, and abilities required for AI security practitioners through workshops and job task analyses. Later phases will include exam development, psychometric validation, and a pilot exam before the certification is made generally available.
France said ISC2 expects the pilot exam to launch in late 2026.
Although ISC2 has incorporated AI-related content into several of its existing certifications, the organization said industry feedback indicated that AI security has become broad enough to warrant a standalone credential.
While ISC2 has not released a final exam blueprint, the development process is expected to examine topics including AI governance, risk management, secure AI development, protection of training data, model security, and emerging attack techniques such as prompt injection, model manipulation, and data poisoning.
The organization is seeking volunteers from a range of cybersecurity disciplines, including practitioners working in AI security, cloud security, governance, software development, security architecture, risk management, and related fields. Participants may contribute to workshops, surveys, exam writing, or pilot testing depending on their experience.
The announcement comes as organizations continue to expand AI deployments across business operations while facing increasing pressure to secure AI models, AI-powered applications, and agentic systems. At the same time, employers have begun seeking more specialized AI security expertise, creating demand for certifications that distinguish practitioners with those skills.
If adopted broadly, the new certification would become one of the first widely recognized, vendor-neutral credentials focused specifically on AI security, joining a growing number of industry efforts to standardize knowledge around securing AI-enabled technologies.
