A structured model for understanding AI risk

The EISA Framework is built around ten pillars. Each pillar represents a specific way risk develops through human interaction with AI.

Rather than broad principles, it breaks risk into practical, observable behaviours. This allows organisations to move from general awareness to structured assessment.

The 10 Pillars

Each pillar targets a specific failure pattern already seen in business environments, from misplaced trust through to loss of accountability.

Orientation - Business Context

Ensures everyone understands AI as an assistive tool, not an authority, with full human accountability for outcomes.

Psychological Stability & Cognitive Integrity

Protects independent thinking by preventing over-reliance on AI and preserving clear, critical judgement.

Continuous Context Awareness

Maintains alignment between AI use and changing business context, ensuring outputs remain appropriate at every stage.

Humanising Boundaries

Keeps AI firmly positioned as a tool, preventing it from being treated as a decision-maker, adviser, or emotional substitute.

Reflection Prompts

Introduces deliberate pauses to verify, question, and validate AI outputs before acting on them.

Leadership Accountability

Ensures the organisation takes ownership of how AI is selected, deployed, and governed, rather than deferring responsibility.

Public Education

Builds shared understanding of AI capabilities and limits so staff and stakeholders use it responsibly and confidently.

Identity Simulation and Authenticity Safeguards

Prevents misuse of AI in ways that imitate people or blur authorship, protecting trust and authenticity.

Adverse Behaviour and Dependency Awareness

Monitors and manages patterns of overuse or reliance to maintain capability, accountability, and healthy behaviour.

Transparency of Collaboration

Makes AI involvement visible so authorship, responsibility, and decision ownership are always clear.

How it works in practice

The framework is designed to be used, not just read.

Organisations assess how AI is being used across teams and workflows, identify where risk is developing, and apply a structured scoring model based on likelihood and impact. This creates a clear view of exposure and prioritises where action is required.

Instead of a static policy, it becomes a working model that supports ongoing oversight.

From framework to behaviour

What makes this effective is that it translates directly into how people work.

It helps teams recognise when AI is appropriate, when it is not, and when additional scrutiny is required. It introduces pause points, reinforces verification, and ensures that responsibility remains visible at every stage of decision making.

Connection to the toolkit

The framework becomes operational through the AI Policy Toolkit. The toolkit takes each pillar and turns it into something practical. It provides structured prompts, assessment tools and guidance that can be applied immediately within a business setting.

Without that structure, risk remains abstract. With it, it becomes something that can be seen, measured and managed.

Buy Now

Limited Offer

For a limited time, the toolkit is available at $850 + GST, including 30 days of direct email support and a 90 day follow up check-in

Purchase and download