AI must not be protected by the right to freedom of expression

The question is not just a rhetorical, parlor exercise, but has relevant consequences. This was raised this week by the Columbia Journalist Review and an expert in law and information explained that the debate is not concluded, but that she considered that no, that generative artificial intelligences do not create "expression", because there is no communicative intention. I agree. Chats only provide a mere reflexive response – however complex – where the algorithm plays with probabilities to satisfy what the user expects to receive. There have already been two concrete cases in the United States where this scrutiny of AI has been tested. First, with that chatbot that encouraged a teenager to commit suicide. Second, with a response from Chat GPT that is considered to have been instrumental in preparing a mass shooting. The first case was resolved with an out-of-court settlement and the second is just beginning to be investigated. Technology companies, needless to say, are asking for their applications to be protected by the First Amendment of the Constitution, where freedom of expression is fiercely guaranteed.

The crux of the matter is understanding that words with meaning do not necessarily constitute discourse. Europe is leading the way: a German court convicted Google for misinformation contained in the AI-generated summary that now appears when you do a search. According to the judge, that text "was not the expression of an acquired conviction of the people expressing it, but the result of an algorithm". The big-tech companies will invest many millions – those they save by parasitizing content without paying for it – in lawyers and lobbyists to win this legal battle. It is, once again, a matter of deceptively invoking freedom of expression to achieve what they are truly pursuing: the right to irresponsibility.