Hello @crypto.piotr,
I am a tad late on this one. It is indeed an issue that had to be aired somewhere. Look the fake news issue is real and has existed for some time now. Let's not go back to 1947:) but let's focus from a period where we felt digital economy had started to take roots.

One example is whatsapp. There are innumerable fake posts which appear on my cell every time and this is all manually sent by someone, we are not even talking automated, leave alone AI.

The root of the issue is our own habit of believing whatever we see as news to be true. The inquisitiveness or reason to check the news is absolutely numb because that is how we have been conditioned since childhood. Our parents, grand-parents watched the news and believed it and we followed suit.

Now, welcome to the 21st century and the era of AI. Frankly, I sell call-center solutions which utilize AI. The respondent is not a person but a bot and uses technology from Amazon Alexa and Google Assistant. So, it is only a matter of time that AI will sound exactly like us.

The question now is not to test whether we will be able to differentiate fake from real, rather test if we can remain on the good side of any subject. Like the Obama example, if a leader is calling us to do something which tramples our conscience then we need to get into the habit of checking those kind of news to be true or not. This is a behavioral change and may take time but that is what is needed.

Lastly, I really can't refute the Cambridge Analytica case, as I am from the country which has been the guinea pigs for them. :)

But here's the thing, the future of AI is either the Terminators (like you said) or the intelligent space scanning interface, as in Star Trek. I personally feel we are headed more towards the Star Trek kind of AI future and hence, the future is indeed bright.