diff --git a/docs/data_access.md b/docs/data_access.md
index 5c9cd665..8a3126b1 100644
--- a/docs/data_access.md
+++ b/docs/data_access.md
@@ -2,13 +2,8 @@
 
 ## Data Download
 
-Please contact greg@mlcommons.org to get access to the data. You will
-need to provide an email for us to add to an access control list.
+Please get the dataset from an official mirror: https://huggingface.co/datasets/MLCommons/peoples_speech
 
-Once you have access, please download and install
-[gsutil](https://cloud.google.com/storage/docs/gsutil). You will need
-to run `gsutil auth login` to log into the same account you provided
-to greg@mlcommons.org.
 
 Then run the following commands:
 
@@ -83,4 +78,4 @@ We show an example script to convert the dataset into a format usable
 by NVIDIA NeMo here:
 [process_peoples_speech_data.py](/scripts/peoples_speech/process_peoples_speech_data.py). NeMo's
 speech recognition input format is described
-[here](https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/stable/asr/datasets.html#preparing-custom-asr-data).
\ No newline at end of file
+[here](https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/stable/asr/datasets.html#preparing-custom-asr-data).