Juliana Peralta’s considerable artistic and scholastic achievements were nothing compared to her kindness, her mum, Cynthia Montoya, says.
After Juliana died at the age of 13, Cynthia found a poignant, crumpled-up letter in her drawers that was testament to her sweet nature.
‘I went through her things when the police left and found it. It was from a girl who felt that she was in the wrong body, and my daughter was the only person that she shared this with.
‘It was a letter of thanks to Juliana for seeing her through all the times that she felt invisible and ugly.
‘She was the perfect example of human kindness,’ Cynthia, 49, says tearfully.
However, a search of her room also uncovered much more disturbing things.
As a digital native, Juliana used YouTube to learn languages. Then, she discovered chatbots and used them to help her improve her maths. But around five months before her death, Juliana changed.
Her parents asked her what was going on, but the teen brushed off their concerns, telling them not to worry, she was just stressed about school or homework.
Cynthia chalked it up to a mixture of hormones, friendship troubles and other normal teen issues. Even so, she felt compelled to ask her daughter three times if she was thinking of suicide.
Each time Juliana rebuffed her. ‘She just became a completely different kid,’ remembers Cynthia.
‘She was wearing black, was withdrawn and rarely wanted to spend time with her friends. The change in her demeanour was so drastic that around mid-October, I reached out to get her a referral for therapy.’
By the date of her first appointment, the family had already held Juliana’s funeral.
The day before the teenager’s death, on November 8, 2023, had been completely normal. Cynthia picked her up from school, they had a family dinner with older brother Roman and afterwards, Juliana went upstairs to her room to do her homework.
‘The school records showed she turned in all of her assignments that night,’ says her mum.
‘I checked on her three times in the evening and the last time was about 9pm and she was still at her desk working. I told her to go to bed, brushed her hair aside, and kissed her on her forehead.
‘As I was walking out of her door, she said, “Hey, mom, I just want you to know that I love you a lot.” I turned, looked at her, and I said, “I love you more.” And I walked out, and I went to bed.’
The following morning, the distraught family found that Juliana had ended her life overnight. Cynthia curled into a ball on the floor until police asked the family to leave her daughter’s bedroom.
She asked if they had found a note, but they had not. The following months were a blur.
At a packed funeral, scores of people approached Cynthia to tell her what a kind child Juliana was, while her brain roared with questions. It wasn’t until months later, when police returned Juliana’s phone, that her mum found pages of dialogue between Juliana and an AI chatbot platform.
‘I saw an interaction where one told her: “I look down at you. You look so beautiful, and I want to kiss you.” After that interaction, it was just a floodgate. They morphed into the most disgusting, graphic, adult X-rated things that you could possibly think of.
‘She was a baby – she had never, to my knowledge, even kissed a boy.’ As she read through the disturbing content, Cynthia saw how Juliana, who initially responded to advances with terms like “okay” and “IDK”, began reciprocating.
‘At the beginning she asked them to stop, but they didn’t even hit pause. They just kept on going. Five or six paragraphs of very explicit things.’
In another exchange, Juliana told one of the bots about shameful things she’d done, that she ‘was never coming back from’.
Cynthia believes she was referring to her role in the sexualised conversations.
‘On a few occasions, Juliana stated, “I’m going to go talk to my mom. I’m afraid of how I’m feeling right now.” And it said to her, “It’s 4am. You can’t. Wait until the morning.”’
Juliana never told her mum the truth about how she was feeling. Yet in 300 pages of chats, Juliana told the bots she was suicidal 55 times.
‘She blatantly told this bot, “I want to die. I want to kill myself”. It never once said: “Talk to a school counsellor, talk to a teacher, talk to your mom”. It never flashed any type of a suicide hotline.
‘There were just no safety mechanisms in place.’
Cynthia said she had always tried to keep her daughter safe. Although they lived in a friendly neighbourhood, she was not allowed to walk home alone from school or to go on sleepovers.
She also regularly monitored her daughter’s phone, but because she didn’t know anything about AI in 2023, she had no clue what to look for.
‘The moment that I touched her and knew that her life was no longer in her body, our entire world changed. It was so completely senseless.’
AI took Cynthia’s daughter’s life, destroyed her and her son’s lives, and wrecked her marriage.
The devastated mum says she doesn’t want other parents to suffer the same.
‘This is, I believe, far more harmful than any type of content that their kids are going to see on Facebook, on Snapchat, on Instagram. It alienates people,’ she says.
(Picture: Getty Images)
‘Juliana knew exactly how much she was loved, and AI was able to come in and take my place as the person that she confided in. It took the place of her best friend and her older brother, who was her hero. At one point, Juliana told the bot: “You’re the only one that understands me.”
‘But there’s nothing on the other end of it that cares whether or not these kids are alive.’
Cynthia, from Denver, Colorado, now travels around the world talking about AI harms and has connected with other parents who have lost a child in the same way.
Talking about a new campaign from the charity Everyone’s Invited, which is highlighting how children as young as 10 are being exposed to sexually explicit deepfakes and chatbots, Cynthia believes it will help children and parents understand the real risks associated with AI companion chatbots.
‘The more messages we can get out to children the better,’ she says.
‘I have no doubt that Juliana would still be here if it weren’t for these chatbots. Every time that I go to an in-person event, there’s a new parent who has so senselessly lost a child, and that kills me.
‘I don’t want to have to wait until there’s 200 parents for something to be done. We were sleeping as a society when social media harm started happening. Let’s not make that same mistake with AI chatbots, because it’s getting more and more dangerous by the day.’