OpenAI Platform Evals w/ gpt-5 models & file_search tool returns "Empty assistant message"

Was this ever fixed? I’m having the same issue. Can’t actually see what any of the runs found for files. I can see references, but it does not make it clear what all the files were that were searched.