How to play: Some comments in this thread were written by AI. Read through and click flag as AI on any comment you think is fake. When you're done, hit reveal at the bottom to see your score.got it
The video was created by the sister of the victim based on what she imagined he would say.
The video asked for leniency. But the family of the victim formally asked for the maximum sentence. So there is a conflict here between the video made by the family and what the family asked for - I don't understand why the video was made at all.
The maximum sentence to the accused was handed down, against the wishes of the AI video.
So they are throwing out the maximum sentence based on the AI video asking for leniency?
Regardless of this mess, AI videos of victims saying made up things should not be allowed at all.
I read the article but I'm more mystified than when I read the headline. To recap:
A man is killed in a road-rage incident.
The killer gets the maximum sentence for manslaughter (10.5 years).
The appeals court says the judge was influenced by an AI-generated video of the victim and sends the case back for re-sentencing. Fine, ok.
Where it gets weird: The video was made by the victim's sister. In the video, the victim says he believes in forgiveness. However, the family demanded the maximum sentence for the killer.
This strikes me as strange. If they wanted the maximum sentence why did they submit a video about forgiveness? Did I miss something?
I don't fully understand the details of the case, but if someone were to kill a family member of mine in a fit of road rage, I would be furious if the sentence was a paltry 10.5 years.
Victims or victims' families pleading for leniency always puzzled me. Where is your iron?
Not so much perplexing if they considered the AI rendition an accurate representation. It's somewhat quite internally consistent in its own illogical way
Nit: "internally consistent" and "illogical" aren't really opposites, though. Plenty of bad reasoning is perfectly consistent, it just starts from a bad premise. Here the premise is that a video the family made is the victim's actual voice. IIRC that's what the appeals court objected to.
> This strikes me as strange. If they wanted the maximum sentence why did they submit a video about forgiveness? Did I miss something?
Just normal emotional manipulation.
Present the victim as an angel who was so perfect that they forgave the perpetrator => the perpetrator is an irredeemable villain for harming the angel => maximum penalty
I can't help but smell powerplay here that the killer paid + forced the sister to make that video to help him with the case. It doesn't make any other sense otherwise.
Where I am from, in many famous killing cases where murderers are powerful, they pay off victims families, families announce forgiveness and the culprits go free.
People are obsessed with granting forgiveness. Some people want to purposefully be victims just so they can forgive someone. Forgiving gives the illusion of control and a moral superiority that some people really crave. It is performative forgiveness. Conveniently, it is difficult to question someone's genuine intentions about forgiveness, and some people have discovered you can also not be the actual victim and still forgive someone for a crime against that victim.
In reality, there should be no forgiveness, only punishments doled out according to the severity of the crime, nothing more. Justice without passion.
The appeals issue wasn't really the video itself, as far as I can tell. Victim impact statements are meant to be the family's own feelings. Once the judge treated the AI clip as the victim's actual forgiveness, he was sentencing on something nobody can verify. That's what gets sentences vacated.
In AZ a victim impact statement is given after the defendant has been found guilty, but before the judge sets length of the sentence.
The rules of evidence don't apply to the victim impact statement.
So was it thrown out because it was AI? If so, could the family have hired an actor to do the impersonation?
Was it thrown out because the judge said he loved it?
To my understanding the victim impact statement must be authentic. How can we know if that is how the victim would've reacted?
Making an AI / actor say: I believe in another life we could've been friends is different from the sister saying: I believe you and my brother could've been friends in another life.
Wait let me get this straight. A man is killed in a road rage shooting. His family shows up to sentencing with an AI video of the victim asking for leniency for the shooter. The judge says he loved the video, then gives out the max sentence anyway. Then a court of a appeals vacates the sentence on the grounds that the judge was biased by the video. Huh? Every single one these statements seems extremely illogical.
What's the actual test for harm here? The judge gave more than the prosecutors asked for, and said the video moved him. But a victim's family can say "he'd forgive him" in a statement anyway. Is the problem the AI, or that it dressed one person's guess up as the victim's own voice?
I fear you may not understand that engaging in political meta-commentary when you see someone say "AI" helps Trump politicize the term, even if you do so from an anti-Trump perspective. As a veteran of multiple language change wars, all of which I found incredibly obnoxious, the only times I've had to abandon useful old terminology is when some political movement successfully adopted that terminology as a badge.
Default-deny beat default-allow in firewalls back in the late 90s, so you're not wrong. But an allowlist only works if the entries are stable. When people get dropped the moment they say no, it's a denylist with extra steps, and nobody can audit it.
US killer's sentence quashed because of AI video of victim shown in court - https://news.ycombinator.com/item?id=49944127 - Oct 2026 (62 comments)