Risk index

The calculator uses the numbers you enter. With 4 harm, 4 autonomy, 3 data, 4 scale, 2 reversibility the result is Result index: 70 / 100. This is not an AI Act class.

This is not an AI Act class. API exposure is on API exposure.

Input data

Results

Enter data and click Calculate.

How it works

Risk index in this calculator is four harm weights plus inverted reversibility. With 4 harm, 4 autonomy, 3 data, 4 scale, 2 reversibility the result is Result index: 70 / 100.

score = 0.19×(sev+aut+dat+sca) + 0.24×(1 − rev), each map15. The formula uses autonomy and reversibility.

Result index: 70 / 100 is not an AI Act class. Another set of numbers gives Result index: 25 / 100, and a further set gives Result index: 95 / 100. Empty optional fields do not change the result.

API exposure answers a different question. Fairness gap does too. Here 4 harm, 4 autonomy, 3 data, 4 scale, 2 reversibility leave Result index: 70 / 100.

Type 4 harm, 4 autonomy, 3 data, 4 scale, 2 reversibility and click Calculate. Two more ready sets sit under the form: Result index: 25 / 100 and Result index: 95 / 100.

Other numbers on the same labels move the result. The first example gives Result index: 70 / 100.

Formula

score = 0.19×(sev+aut+dat+sca) + 0.24×(1 − rev), each map15.

How to use

  1. Type 4 harm, 4 autonomy, 3 data, 4 scale, 2 reversibility.
  2. Leave optional fields empty if you do not want them in the result.
  3. Click Calculate. This set shows Result index: 70 / 100.
  4. The second example under the result is Result index: 25 / 100, the third Result index: 95 / 100.
  5. This is not an AI Act class.

Risk index = Result index: 70 / 100

The calculator uses the numbers you enter. With 4 harm, 4 autonomy, 3 data, 4 scale, 2 reversibility the result is Result index: 70 / 100. This is not an AI Act class.

Risk index
Risk index is this calculator's result. The first example is Result index: 70 / 100.
autonomy
The formula includes autonomy. The second example gives Result index: 25 / 100.
reversibility
The formula includes reversibility. The second example gives Result index: 25 / 100.

Examples

Example 1

  • 4 harm
  • 4 autonomy
  • 3 data
  • 4 scale
  • 2 reversibility

Result index: 70 / 100

The calculator uses the numbers you enter. With 4 harm, 4 autonomy, 3 data, 4 scale, 2 reversibility the result is Result index: 70 / 100. This is not an AI Act class.

Example 2

  • 2 harm
  • 2 autonomy
  • 2 data
  • 2 scale
  • 4 reversibility

Result index: 25 / 100

With 2 harm, 2 autonomy, 2 data, 2 scale, 4 reversibility the result is Result index: 25 / 100.

Example 3

  • 5 harm
  • 5 autonomy
  • 5 data
  • 4 scale
  • 1 reversibility

Result index: 95 / 100

With 5 harm, 5 autonomy, 5 data, 4 scale, 1 reversibility the result is Result index: 95 / 100.

Related calculators

API exposure

API exposure: score = mean(map15) × 100.

Use calculator →

Fairness gap

Fairness gap: gap = |group A − group B|.

Use calculator →

All AI regulation calculators

Common questions

What do you get from 4 harm, 4 autonomy, 3 data, 4 scale, 2 reversibility?

Result index: 70 / 100. This is not an AI Act class.

What about the second example?

Result index: 25 / 100.

What about the third example?

Result index: 95 / 100.

Is Result index: 70 / 100 an AI Act class?

No. It is the result from the numbers you enter.

Do empty fields change Result index: 70 / 100?

No. Empty optional fields do not change the result.

How is this different from API exposure?

That calculator answers another question. Here Risk index stays at Result index: 70 / 100.

Does a comma in the number work?

Yes. A comma and a dot mean the same in a number field.

Is the result a diagnosis or official opinion?

No. The result comes from the numbers you enter. an AI Act class sits outside this calculator.

Do 2 harm, 2 autonomy, 2 data, 2 scale, 4 reversibility give Result index: 25 / 100?

Yes. That is the second example under the form.

How is the risk index calculated?

The calculator maps five 1 to 5 scores onto 0 to 1. Then it takes 0.19 times the sum of harm, autonomy, data, and scale, plus 0.24 times (1 minus reversibility). In the first example the result is Result index: 70 / 100.

Why is reversibility inverted?

Higher reversibility lowers the index: the formula takes 1 minus mapped reversibility and multiplies by 0.24. The other four scores raise the index.

Is this an AI Act class?

No. The result is a heuristic from the five scores you enter. The calculator does not assign a class under the AI Act.

Can I change the 0.19 and 0.24 weights?

No. The weights are fixed in this page formula. The fields only take scores from 1 to 5.

Knowledge sources

The terms come from EU acts. The calculator result is a sketch index, not a regulator decision.

Page updated in 2026.