Academic researchers published attack related to Google's "intelligence"

Academic researchers have published an attack that led Google's supposed "intelligence" to obey malicious commands to manipulate devices in the user's home.

Giving Google control of your devices, or control of your own computing that you do on their servers, inevitably makes you vulnerable to Google. This announcement shows that the vulnerability includes third-party crackers too.

Please don't call these systems "artificial intelligence"; >intelligence is something they do not have.

Side point: The article says that the crack discoverers worked with Google to "mitigate" the danger. What, concretely, does "mitigate" mean here? I think in this case it is a weasel word to suggest fixing a problem without claiming to have fixed it.

</li>