It's not lying. Maybe it's being wrong, maybe it's negligence, but it's not lying. It's not like they're trying to display the wrong information. The algorithm isn't a person. It's not as simple as saying "oh, our algorithm messed up, we'll kindly tell it to say that you live over there from now on".

I wouldn't call it lying, I'd call it a sub-optimal algorithm design.

Generally, the person or group of people behind an algorithm needs to be fully accountable and responsible for the algorithm they use to interact with the world.

If saying: "it's not me, it's the algorithm" gets you out of the responsibility you can do all kinds of morally questionable things.

(edit: grammeer)

"It's not me, it's the killbot algorithm I designed, patented, and licensed!"

In seriousness, though, I agree and we're already way behind the eight-ball on this one, given predictive policing and bail algorithms.

Wikipedia and Buzzfeed and other websites suffer from being wrong as well. Trying to get a bias in the news to get more clicks that might be wrong.

At what point does it become fake news or just being wrong?

To me, it's about intent. If they are unknowingly wrong, they aren't lying. As soon as somebody informs them of the error, if they continue to report falsehoods, they are lying (and have now become fake news). It doesn't matter if the falsehood is generated by an algorithm or a person, or if the falsehood is malicious or not - as soon as they know they are wrong, they should take action to correct whatever has been published.