The ASD Essentials is technology-agnostic, so it applies whether an organisation consumes, develops or deploys AI. ASD’s AI considerations guidance explains how the existing principles, actions and domains apply to AI, framing it three ways at once: as a source of cyber risk, as a capability that can support defence, and as an asset that must be identified and governed.
Does AI change how the ASD Essentials works?
No. The framework is technology-agnostic and focuses on security outcomes that apply to AI-enabled and traditional systems alike. ASD is clear that the ASD Essentials applies regardless of whether an organisation consumes, develops or deploys AI, and that most principles apply to AI the same way they apply to other technology, even if the implementation differs. The AI considerations guidance does not add a separate AI framework; it explains how to scope, implement and assess the existing principles and actions when AI is involved, with notes for the enterprise IT, cloud and OT chapters. This mirrors the approach in the new ISO/IEC 27090 guidance, which adds AI specific threats on top of established security rather than replacing them.
How does the ASD Essentials treat AI as a threat?
As a force multiplier for attackers and a new attack surface. ASD’s position is that AI has not fundamentally changed how adversaries gain access and move through an environment, but it has made existing attacks faster, cheaper and available to more people, which weakens assumptions organisations have relied on for years. AI helps adversaries run reconnaissance, find vulnerabilities and build exploits or malware more quickly, and synthetic voice and video can impersonate trusted people to deceive help desks or those who approve payments or access. AI systems are also targets: indirect prompt injection through emails, documents or web content processed by AI agents, tampered models pulled from public sources, and sensitive data sent to unapproved AI services. ASD’s answer is layered protection, so defeating one safeguard meets another.
How can AI help defence, and what is the catch?
AI can speed up defence, but accountability stays with people. AI-enabled security tools can process large volumes of data, correlate events across systems, identify vulnerabilities, triage alerts and synthesise threat intelligence, helping defenders respond faster and spend human effort where judgement is needed. The catch is governance: an AI-enabled security product is itself an AI asset and should be assessed, configured and monitored like one. ASD advises using these tools to support the framework’s actions rather than replace them, and validating AI-generated outputs before they inform significant decisions. In other words, AI assists analysis and decisions, but the organisation remains accountable for the outcome. This is the same caution behind shadow AI risks.
How are AI systems treated as assets?
As in-scope assets that must be identified and governed like any other. The Essentials treats AI systems as assets within scope, which can include internally developed or fine-tuned models, the infrastructure that hosts or trains them, external AI services, AI-enabled platforms and applications, training and fine-tuning data, prompts, outputs and inferences, interfaces, connectors and APIs, and the accounts and non-person identities used by AI agents and services. AI tools can hold identities, permissions and access of their own, and should be governed like other identities. When scoping, an organisation weighs an AI asset’s function, connectivity, criticality, access and the consequence of its compromise, then the relevant chapter determines how the principles and actions apply. In the enterprise IT chapter, for example, a public-facing AI service maps to boundary services, an AI agent’s identity to identity services, and an AI tool with administrative reach to privileged environments.
Who is accountable for AI at the top?
The board or executive committee, through an explicit ISM principle. The Essentials draws on the Information Security Manual’s executive artificial intelligence accountability principle, under which the board or executive committee is accountable for ensuring AI is secure, controllable, human-supervised and used ethically and accountably. In practice that means establishing executive accountability for the use, governance and risk management of AI, supported by processes for approval, risk assessment, oversight and ongoing review of AI use cases. This connects the technical controls to board-level governance, and lines up with the 2025-26 threat report naming AI as the defining force in the threat environment. Cybernion helps organisations govern and secure AI through AI governance readiness.
Frequently asked questions
Does the ASD Essentials have special AI rules?
Not separate rules. The framework is technology-agnostic and applies to AI-enabled and traditional systems alike. ASD’s AI considerations guidance explains how the existing principles, actions and domains apply to AI.
How does the ASD Essentials frame AI?
Three ways at once: AI as a source of cyber risk, AI as a capability that can support defence, and AI as an asset that must be identified and governed like any other.
Who is accountable for AI security under the Essentials?
The board or executive committee, through the ISM’s executive AI accountability principle, which makes leadership accountable for AI being secure, controllable, human-supervised, ethical and accountable.
Written by Gaurav Vikash, an ASD endorsed IRAP assessor and senior cyber security leader with 18 years of experience across Australia, the UK and Asia, including CISO and senior security leadership roles. He holds CISSP, CISA, CISM and CRISC and is an ISO 27001 Lead Auditor and ISO 42001 Lead Implementer, and speaks regularly at industry conferences.
If you are adopting AI and want it governed under the Essentials, talk to us.
Sources:
- AI considerations within the Essentials, cyber.gov.au, 15 October 2026
- ASD Essentials, cyber.gov.au, October 2026
- ISO/IEC 27090:2026, iso.org, October 2026
Last updated: 27 October, 2026
