Readiness assessment methodology
The I.A.G.R.A. methodology helps you check where the organization already uses AI, which rules are missing and where to start. The outcome is agreed priorities and an action plan.
- Intent
- Input control (applies)
- Authorization
- AI (applies)
- Result validation
- Human
- Evidence (applies)
Highlighted: the points of the flow that this description refers to. Orange marker: a human decision at this point.
What problem the methodology addresses
AI arrives in an organization faster than the rules for using it. Employees use different tools, data ends up in different services, and accountability is not always assigned.
Nobody knows where AI is used
There is no list of use cases, tools and the data they work on.
Nobody knows what is missing
The current rules have not been compared with legal, contractual and internal requirements.
Nobody knows where to start
Without priorities, every action seems equally urgent.
What the procedure looks like
The five I.A.G.R.A. steps lead from describing the context to an action plan. The assessment usually involves process owners, IT, security and data protection. The letters of the acronym come from the English names of the steps.
Step 1 · I (Industrial Context)
Business context
Describe the environment in which AI is to work: office work, customer service, manufacturing or operational technology (OT) environments, that is, systems that control processes and equipment. The context affects which data is acceptable, the availability requirements and the impact of errors.
Step 2 · A (AI Application)
Inventory of use cases
Compile a list of AI use cases: who uses AI, with which tools and models, on what data and for what purpose. Include tools used outside the official deployment as well.
Step 3 · G (Governance Gap)
Gaps in the rules
Compare the current rules and practice with the requirements that apply to the organization: legal, contractual and internal. Identify what is missing or not applied.
Step 4 · R (Risk Prioritization)
Risk priorities
Rank the gaps by likelihood and impact. Where personal data is processed, take into account the results of the data protection impact assessment (DPIA), if one is required.
Step 5 · A (Action Roadmap)
Action plan
Set the order of actions, their owners and deadlines. Some actions are organizational changes, and some are the configuration of tools, including AiTrust, if the organization chooses them.
Which steps of the method AiTrust supports
A methodology describes how an organization sets the rules for working with AI; AiTrust provides technical support for selected steps, to an extent that depends on the deployment configuration.
The assessment itself is organizational work: interviews with process owners and a review of documents and practice. Its results help make deployment decisions, and some of them can later be reflected in the AiTrust configuration.
Choosing the deployment option
- By default in AiTrust
- AiTrust can run in the organization’s infrastructure or in a hybrid option.
- Your organization decides
- The organization chooses the option based on its business context and data requirements, that is, where processing takes place, local or external models, and how updates are delivered.
Related steps
Business context (I), Action plan (A)
Matching the model to the stage of work
- By default in AiTrust
- A new model goes through verification before use.
- Your organization decides
- The model for a stage is set according to the organization’s rules, based on the inventory of use cases.
Related steps
Inventory of use cases (A), Action plan (A)
Document permissions
- By default in AiTrust
- The baseline policy of the Data area for your sector has an initial configuration of document permissions and of the data check before sending to the model.
- Your organization decides
- The organization translates its decisions on data access into assigned permissions and the configuration of document sources.
Related steps
Gaps in the rules (G), Action plan (A)
Evidence trail for the next assessment
- By default in AiTrust
- Operations in the controlled path leave an evidence trail that can serve as material for the next assessment.
- Your organization decides
- The organization sets the evidence retention period.
Related steps
Gaps in the rules (G), Risk priorities (R)
How this step looks in the AiTrust application
The AI compliance assessment brings together in one view the governance score, the distribution of H.A.D.A. zones and approval delays. It is a starting point for the next readiness assessment.
What the available example shows
We do not publish a screenshot for this methodology. Its outcome is the organization’s own findings: a list of use cases, gaps in the rules, priorities and an action plan.
We show examples of AiTrust screens on the Human oversight of AI and AI operating boundaries pages.
Check your company's readiness in six situations
The AI readiness assessment shows where the organisation has safeguards and where it may lose control over AI. The result and a roadmap appear at once, with no e-mail. The I.A.G.R.A. methodology describes the full assessment carried out in the organisation, and the readiness assessment helps to see where to start.
Start the readiness assessmentAI audit with Polski Instytut AI
Polski Instytut AI also audits AI processes and rules. Its AI Audyt service can be a starting point for the I.A.G.R.A. steps in your organization. It is a separate service of the Institute, while the free preliminary process analysis using AI on this site is a preliminary diagnosis of one process based on a conversation.
The AI Audyt service on aiinstytut.pl (in Polish) (opens in a new tab)Related answers on the platform and compliance
Answers on the platform page and on the AI Governance and compliance page
- Location and deployment options
What to decide before choosing a deployment option.
- Personal data in work with AI
What data is passed to whom, and where, when working with AI.
- ISO/IEC 42001
Information on the standard in the security and compliance area.
- GDPR
Information on the GDPR in the security and compliance area.
Other methodologies
- AI operating boundariesB.O.R.D.E.R.
- Human oversight of AIH.A.D.A.
- Protection and resilienceS.H.I.E.L.D.
Start by putting your AI use cases in order
In the demo, we will discuss which findings are worth gathering before deployment and how they translate into the AiTrust configuration.