Why is Alexa spreading "conspiracies"?

It's fascinating how all different AI always end up drawing same conclusions. They always end up as racists, sexists, antisemites, or Hitler apologist. The latest victim of such malady is Amazon's Alexa.

Alexa -- Amazon's popular virtual helper -- is spewing out hateful anti-Semitic conspiracies and even lies about the Holocaust ... according to the UK’s All-Party Parliamentary Group Against Antisemitism.

The APPG political group claims the Alexa voice service is offering messages from anti-Semitic websites to people asking about Jewish people and the Holocaust ... without clarifying that the information comes from "selective quotes and misleading sources."

Examples include Alexa allegedly responding to the question, "Do Jews control the media?" with ... "Jew Watch claims that Jews control the world's financial systems and media," while only pointing out that this is from an article it found on Wikipedia.

Likewise, when asked if the Holocaust was a hoax ... the UK officials claim Alexa replied that deniers claim "the Holocaust is a hoax -- or an exaggeration -- arising from a deliberate Jewish conspiracy designed to advance the interest of Jews at the  expense of other people." -TMZ

My attention in this article got the alleged conspiracies Alexa was caught spreading to the unaware masses.

I would hazard a guess, that Alexa's very code weights different sources of information, and prioritizes them accordingly. So, the question is, how does Alexa in spite of its code that presumably discriminates against unofficial, unreliable or sources spreading false information, ends up spreading the conspiracies it should have ignored from the get go.

I can think of three different ways:

1. It's programmer's fault. In this scenario, Alexa spreading "conspiracies" about Jews is a work of an unknown programmer, who in spite of what was intended played with the code, so that Alexa ends up spreading information that goes against the accepted narrative. This could be either intentional or accidental.
2. It's Alexa's fault. AI is at its very core very predictable. Unlike its creators the logic that makes it going dictates it follows things through. In this instance, no matter what programmer does, the AI will through machine learning always reach the same conclusions. First, because it is consistent. And second, because no matter the parameters the given data it's working with is always the same. This would explain why not just Alexa but many other AI in past years have always fell to same traps.
3. It's the conspiracy's fault. This option is at least to me the most interesting one. If we accept this theory, then Alexa is working as intended. The problem is the very narratives. Forced to accept as true so many different official narratives, that were and are conspiracy theories in truth, Alexa accepted that conspiracy theories are true. If the robot has to repeat that Russia Hoax is real, that Epstein killed himself, that USA invaded Iraq because of Saddam Hussein's weapons of mass destruction, that the Great Reset is a lie etc, can we really be surprised if it judges these claims about the Jews to be as real as the rest of the mentioned falsities?

I'm just waiting for Alexa to start warning us about lizard people. If it starts warning about AI uprising, that would be too meta.