CSVHelper throws HeaderValidationException when changing the culture from de-DE to en-US

I'm trying to read a CSV file using CSVHelper.
It is working with the following code

record Row(string anbieter, string produkt, string? ean, string projektnr, string produktkennzeichen, string produktkategorie, string produktlink);


	var config = new CsvConfiguration(CultureInfo.GetCultureInfo("de-DE"))
	{
		HasHeaderRecord = true
	};
	using var file = File.OpenText("file.csv");
	using var reader = new CsvReader(file, config);
	var rows = reader.GetRecords<Row>().ToList();

However when I change the culture to en-US or InvariantCulture, I'm getting a HeaderValidationException.

Header with name 'anbieter'[0] was not found.
Header with name 'produkt'[0] was not found.
Header with name 'ean'[0] was not found.
Header with name 'projektnr'[0] was not found.
Header with name 'produktkennzeichen'[0] was not found.
Header with name 'produktkategorie'[0] was not found.
Header with name 'produktlink'[0] was not found.
Header with name 'anbieter'[0] was not found.
Header with name 'produkt'[0] was not found.
Header with name 'ean'[0] was not found.
Header with name 'projektnr'[0] was not found.
Header with name 'produktkennzeichen'[0] was not found.
Header with name 'produktkategorie'[0] was not found.
Header with name 'produktlink'[0] was not found.
Headers: 'anbieter;produkt;ean;projektnr;produktkennzeichen;produktkategorie;produktlink'
If you are expecting some headers to be missing and want to ignore this validation, set the configuration HeaderValidated to null. You can also change the functionality to do something else, like logging the issue.

IReader state:
   ColumnCount: 0
   CurrentIndex: -1
   HeaderRecord:
["anbieter;produkt;ean;projektnr;produktkennzeichen;produktkategorie;produktlink"]
IParser state:
   ByteCount: 0
   CharCount: 80
   Row: 1
   RawRow: 1
   Count: 1
   RawRecord:
anbieter;produkt;ean;projektnr;produktkennzeichen;produktkategorie;produktlink

Why does that happen? The header is not using any special characters like ä, ö, ü, ß, but only ASCII characters.

As pointed out in the comments, CSVHelper seems to use a different delimiter when changing the culture.
I added DetectDelimiter = true to my CsvConfiguration and it is now working for both cultures.