TL;DR
- NIST released two new CSF 2.0 Quick-Start Guides in early 2026 covering enterprise risk management and implementation pathways for diverse audiences (NIST CSF 2.0 Quick Start Guides).
- The draft NIST Cyber AI Profile (NIST IR 8596) maps AI-specific risks to CSF 2.0 functions, addressing securing AI components, using AI for cyber defense, and countering AI-boosted attacks.
- NIST CSF 2.0 introduced a new Govern function, a sixth core function requiring organizations to establish and monitor cybersecurity governance structures, including for AI (NIST CSF 2.0).
- OWASP maintains a Top 10 list of the most critical risks for LLM applications (OWASP Top 10 for LLM Applications), expanding the cybersecurity and governance considerations organizations need to address when using large language models.
- If you are using NIST CSF 2.0: your GRC program must now account for AI-specific control requirements.
NIST released the draft Cyber AI Profile (NIST IR 8596) in December 2025 and two new CSF 2.0 Quick-Start Guides in early 2026, marking the formal arrival of AI-specific cybersecurity controls in the world's most widely adopted security framework. Organizations that have already implemented CSF 2.0 must now assess how their GRC programs handle AI-specific risks, and those still on CSF 1.1 face a compounding gap.
NIST CSF 2.0, released in February 2024, made two structural changes that GRC teams are still catching up with in 2026:
CSF 2.0 added a sixth core function, Govern, sitting alongside Identify, Protect, Detect, Respond, and Recover. The Govern function requires organizations to establish, communicate, and monitor cybersecurity risk management strategy, expectations, and policy across the enterprise. This is not optional augmentation; it is a core framework function that must be evidenced and monitored.
For GRC programs still mapped to CSF 1.1, the Govern function represents a structural gap. Control mappings, evidence libraries, and assessment questionnaires built for CSF 1.1 do not cover Govern requirements. Organizations that have not updated their GRC frameworks to CSF 2.0 have an unmapped compliance exposure.
NIST’s ongoing expansion of CSF 2.0 informative references, including guidance relevant to finance, manufacturing, telecom, and AI, means organizations in these sectors are seeing increasing specificity around how cybersecurity risk management should be applied in their industry contexts. Meanwhile, OWASP’s LLM Top 10 highlights the growing need for AI-specific security controls and governance practices that organizations can align with broader CSF 2.0 cybersecurity outcomes.
The draft NIST Cyber AI Profile applies CSF 2.0 to three distinct AI cybersecurity focus areas:
1. Securing AI components
AI systems, including models, training pipelines, APIs, agents, and inference infrastructure, introduce novel and expanded attack surfaces that organizations must address within existing cybersecurity programs. The Cyber AI Profile includes guidance related to adversarial inputs, data poisoning, model security, supply chain risk, and protection of AI systems and supporting infrastructure.
2. Conducting AI-enabled cyber defense
Organizations are increasingly using AI to support threat detection, anomaly detection, investigations, and automated response workflows. The Profile addresses governance, oversight, reliability, and operational considerations for AI-enabled defensive capabilities, including human oversight, model drift, and confidence validation.
3. Thwarting AI-enabled cyber attacks
Threat actors are using AI to accelerate phishing, social engineering, malware development, reconnaissance, and attack automation. The Profile focuses on helping organizations build resilience against AI-enabled threats through updated defensive practices, training, detection, and response capabilities.
For organizations already operating a NIST CSF-aligned GRC program, the path to CSF 2.0 and Cyber AI Profile alignment follows four steps:
Without a GRC platform that supports dynamic control library updates and multi-framework mapping, this process typically takes months and generates significant documentation debt.
6clicks maintains an active content library that includes NIST CSF 2.0 controls, including the Govern function, and is tracking the Cyber AI Profile as it moves from draft to final publication. Practical capabilities include:
Always audit-ready means your NIST CSF 2.0 compliance posture is visible in real time — not reconstructed at audit time.
If your GRC program has not yet incorporated the NIST CSF 2.0 Govern function or begun mapping to the Cyber AI Profile, now is the time to act, not after the Profile is finalized. Book a demo of 6clicks to see how automated framework mapping and real-time gap tracking keep your NIST CSF 2.0 program always audit-ready.