This issue is now resolved.

We have identified that the root cause of the issue is with one of our transcription pipelines and have routed transcription to different providers until a full resolution is in place.