---
title: Criteria of the AI Potential Assessment
description: >-
  How the Scoring by AI Potential report calculates: the four criteria AI
  Potential, Frequency, Process Structure and Compliance Relevance, their scores
  and the total weighting.
search:
  tags:
    - nachschlagen
    - ki
    - kuenstliche-intelligenz
    - potenzial
    - bewertung
    - kriterien
---
The **Scoring by AI Potential** report assesses each business process based on four criteria and adds up the scores to give the **Total Weighting** — the higher the value, the greater the expected potential for AI support.

Each criterion is assessed using defined values, each of which is assigned a numeric score.

## Assessed Criteria [#assessed-criteria]

Four criteria are used for the scoring:

- AI Potential
- Frequency of Process Execution
- Process Structure
- Compliance Relevance

The individual criteria and their values are described below.

## AI Potential [#ai-potential]

This criterion assesses the extent to which the process is fundamentally suitable for the use of AI or already uses such support.

| Value | Description | Score |
| --- | --- | --- |
| No AI use | The process currently does not use any AI support. | 75 |
| AI supported | AI can support individual process steps. | 45 |
| AI integrated | AI is already an integral part of the process. | 5 |
| No action required | The process is not suitable for the use of AI or does not require any support. | 0 |

High values arise in particular for processes that do not yet use AI but have great potential for future use.

## Frequency of Process Execution [#frequency-of-process-execution]

This criterion assesses how often a process is executed.

Processes with a high execution frequency often offer greater potential for automation and efficiency.

| Value        | Description                                                  |   Score |
|:-------------|:-------------------------------------------------------------|--------:|
| up to 10 per year  | The process is executed rarely.                        |      15 |
| up to 10 per month | The process is executed regularly, but relatively rarely. |   30 |
| up to 10 per day   | The process is executed frequently.                    |      45 |
| up to 100 per day  | The process is executed very frequently.               |      60 |
| more than 100 per day | The process is executed extremely frequently.       |      75 |

The higher the frequency, the greater the potential efficiency gain through AI support usually is.

## Process Structure [#process-structure]

This criterion assesses the degree to which a process is structured. Flexible processes receive more points here than highly structured ones: where a process frequently requires individual decisions, AI can offer the most support. In the **Scoring by Digitalization Potential** report, it is the other way round.

| Value | Description | Score |
| --- | --- | --- |
| Flexibility required | The process frequently requires individual decisions or adjustments. | 40 |
| Rather flexible | The process has partly flexible workflows. | 30 |
| Rather clear structure | The process largely follows clearly defined workflows. | 20 |
| Clear structure without any scope for deviation | The process is fully standardised. | 10 |

## Compliance Relevance [#compliance-relevance]

This criterion assesses which regulatory or organisational requirements have to be taken into account when executing the process.

| Value          | Description                                                                               |   Score |
|:---------------|:------------------------------------------------------------------------------------------|--------:|
| External       | The process is subject to external regulatory specifications or legal requirements.       |       0 |
| Internal       | The process is subject to internal guidelines or compliance specifications.               |      15 |
| Not relevant   | There are no particular compliance requirements.                                          |      30 |

Processes with no or low compliance relevance are often better suited to AI-supported optimisations.

## Calculation of the Total Score [#calculation-of-the-total-score]

The total score of a process is the sum of the individual assessment values.

**Score = AI Potential + Frequency + Structure + Compliance**

The assessment is carried out automatically based on the field values stored for the process.

In the report, the result is shown in the **Total Weighting** column; the four columns before it show the value for each criterion.

## Prioritisation of the Processes [#prioritisation-of-the-processes]

The prioritisation within the report is based on the calculated score.

- Processes are sorted by descending score.
- Processes with high scores have a particularly high potential for the use of AI.
- These processes can be specifically considered as priority candidates for AI initiatives or automation projects.

See also: [Available Reports](/en/architecture/reports/list-available-reports)