What on Earth is the basis for admissibility of such evidence and/or testimony?

From the original source, statement of a judge identified as "chief justice Timmer" statement:

AI has the potential to create great efficiencies in the justice system and may assist those unschooled in the law to better present their positions. For that reason, we are excited about AI’s potential. But AI can also hinder or even upend justice if inappropriately used. A measured approach is best. Along those lines, the court has formed an AI committee to examine AI use and make recommendations for how best to use it. At bottom, those who use AI—including courts—are responsible for its accuracy.

<https://www.abc15.com/news/region-southeast-valley/chandler/...>

I'm presuming that this is Ann A. Scott Timmer, chief justice of the Arizona state supreme court:

<https://en.wikipedia.org/wiki/Ann_Timmer>

Yeesh. I understand that this kind of thing might have some healing effect on hurt people, but I can't get over the ramifications of treating an AI like a real person in a court setting. For instance, imagine a suspect skips out on a hearing due to some unknown circumstance - in absentia, the court relies on an AI testimony trained on the suspect, who then changes his plea to guilty and confesses to the crime.

If I was killed in something as pointless as a road range incident and then an AI trained on me was wheeled into court to "forgive" my aggressor I'd be incensed from beyond the astral plane.

I'm going to add a section to my will.

It will stipulate that if my family does this with my likeness, they must:

1) key in a slowly gyrating field of lines as the background

2) occasionally add a cd-skipping effect, and occasionally garble the pitch

I'll be gone, so it won't make a difference, but it cracks me up to think of a solemn courtroom suddenly watching Thom Headroom give an impact statement.

Surprised a judge allowed this

Whoa this is black mirror dystopia level stuff!