One research group is training an AI model to flag "antisemitic content" (in the usual extremely strict way).
That's pretty shitty, but there's no affiliation with the German government here, nor anything else that supports a claim of "Germany wants". That's just ragebait.