Azure ML Execute R Script component - security error

Question

Azure ML Execute R Script component - security error

Bill Ehrreich 0

For purposes of testing R functionality with Azure ML, I have created a simple pipeline with an Execute R Script step. No matter what script I put in the step, when the job runs, it ends in an error. The error is shown below in pink at right: "The code of this job is not protected by our systems and not secure...". What do I need to do to get around this error so that I can use the Execute R Script step?

User's image

1 answer

Your answer

Answer 1

Sina Salam 22,576 Volunteer Moderator

Hello Bill Ehrreich,

Welcome to the Microsoft Q&A and thank you for posting your questions here.

I understand that your Azure ML Execute R Script component is having security error.

The error is because you're using the inline R script or unvalidated script bundle which does not meet Azure ML’s security validation requirements. Inline scripts are not validated for security, so avoid it and always use script bundles for:

Production pipelines
Sensitive data processing
External package usage

Therefore,

Use Script Bundle Instead of Inline Code by:
- Creating a .zip file containing your R script (e.g., my_script.R),
- Upload it as a File Dataset in Azure ML Studio,
- Connect it to the Script Bundle input port of the Execute R Script component.
In the link you will see how you can execute R Script Component - https://learn.microsoft.com/en-us/azure/machine-learning/component-reference/execute-r-script?view=azureml-api-2

Always use proper entry point in your script, it must define:

   azureml_main <- function(dataframe1, dataframe2){
     source("./Script Bundle/my_script.R")
     result <- my_func(dataframe1)
     return(list(dataset1=result, dataset2=dataframe2))
   }

In this link you will find AzureML-R-template on GitHub - https://github.com/microsoft/AzureML-R-template

Use CLI or YAML for an advanced R workloads and if designer fails, switch to CLI-based workflows:
- See examples here on how you can achieve it - https://github.com/MicrosoftDocs/azure-ai-docs/blob/main/articles/machine-learning/component-reference/execute-r-script.md
- Define jobs in YAML and deploy using az ml job create

I hope this is helpful! Do not hesitate to let me know if you have any other questions or clarifications.

Please don't forget to close up the thread here by upvoting and accept it as an answer if it is helpful.

Bill Ehrreich 0 Reputation points

2025-08-11T13:19:43.55+00:00

Sina,

Thank you for your response. I have already tried using the script bundle, but it results in the same error. Please see my example below.

Bill
Sina Salam 22,576 Reputation points Volunteer Moderator

2025-08-11T14:14:54.9533333+00:00
Hi Bill Ehrreich,

Thank you for your feedback.

As I have said above:

Try as much as possible to use latest CLI/SDK - run 'az upgrade' and update ml extension and also, azure-ai-ml SDK must be up-to-date.

If all fail, submit via CLI/YAML instead of Designer by:

Define your job in YAML or Python the example YAML snippet will look like :
command: > Rscript entrypoint.R --inputs dataset1=@{inputs.dataset1} --outputs output1=...

Then, submit using:
az ml job create --file job.yml

And should there be any case that the issue persists:

Check Outputs + logs > 70_driver_log for detailed error.

Validate that your ZIP is correct and not causing extraction issues

Thank you for your understanding.
Bill Ehrreich 0 Reputation points

2025-08-11T14:37:03.89+00:00

Sina,

Thank you again for your response. I have the latest CLI installed. I installed it 3 days ago.

What you seem to be saying is that you don't know how to make this work on Azure Machine Learning Studio, so I should instead try to use YAML and the CLI. I have not tried that yet, but it seems silly to provide an execute R script component that does not work in the designer.

Bill
Sina Salam 22,576 Reputation points Volunteer Moderator

2025-08-11T16:38:53.8766667+00:00

Hi Bill Ehrreich,

Thank you for your feedback.

Regarding this statement:

What you seem to be saying is that you don't know how to make this work on Azure Machine Learning Studio, so I should instead try to use YAML and the CLI.

Instead of getting stuck, there are many ways of getting things done and that's an optional practicable for many instead of this common error, not that I don't know. Alternatively, if you can rewrite your function so it not only loads the script but also processes and outputs the modified data. Doing that will help avoid “security error” issues if Azure ML detects unused script execution.

Success
Bill Ehrreich 0 Reputation points

2025-08-11T17:37:10.6666667+00:00

Sina,
I tried to rewrite my script so that the script bundle took care of the entire function, with the inputs and outputs, but that left an empty box for the Execute R Script component, and this fails validation.
Bill

Share via

Azure ML Execute R Script component - security error

1 answer

Your answer