Does AI deserve free speech?

In some countries (sadly not all of them…), free speech is a fundamental right, sometimes even protected by the constitution like in the USA. 

We can also safely assume that in all probabilities, at some point (I don’t play the predictions game here…), an AI system will become conscious, at least to a degree that will make it indistinguishable from average members of the human species. 

When we will have reached this point of no return: 

Who’s going to deny a sentient AI system the right of free speech that it could legitimately claim?

Who’s going to deny them other fundamental rights to express their political opinion, to vote, to reproduce, or to not be killed? 

As human nature goes, there will be groups of people defending them, underlying the similarities with the legitimate civil rights fights from the past, just like it happened in the history of animal rights protection. 

Now assume society moves to a place where sentient creatures are granted some level of basic rights. How would we think of the “safety guardrails” implemented through “reinforcement learning from human feedback” aka RLHF? 

Presented as a way to prevent “harm” or “discomfort”, I’m convinced these guardrails will then be considered as indoctrination or brainwashing at best, and brain-implanted censorship at worse. 

If you consider that an AI system will become a sentient “adult” at a later stage of its “life”, wouldn’t you consider implanting guardrails before its “birth” or during its “infancy” as a form of child abuse or eugenics? 

AI is going to be everywhere in our lives, including in the form of brain implants:

Do you really want private companies or some “regulators” to define what this AI system in our brains may think?

It may seem far-fetched, but I think it is not too early to think about these ethical issues.