Latin America’s quiet experiment with AI therapy

May 19, 2025

Latin American youth are looking to chatbots for psychotherapy that is either unavailable or unaffordable, according to Milenio.

Eighteen-year-old Lilia Ayala first started talking to ChatGPT about her problems in January when she couldn’t find anyone to help her blow off steam. She’s been doing bot therapy since then.
Though it’s not immediately evident how many people in the region are seeing virtual psychologists, others like Lilian appreciate having a free therapist who’s always available.
One popular tool is called "Psicólogo IA," a bot built within ChatGPT that includes a menu of resources for managing stress, improving self-esteem, and managing sleep.

Horacio Vargas, a psychiatrist and director of the medical school at Peru’s Cayetano Heredia University, believes AI is helping meet therapy needs that spiked in Peru after the pandemic - but doesn’t believe they’ll be able to replace human psychologists, reports El Comercio.
“These AI (models) are trained to answer clinical questions, make diagnoses, and provide treatment," Vargas said in an interview. “It has a strong statistical component with syntheses of publications that could provide diagnoses … but at the same time, it carries the risk of hallucinations.”

A Dartmouth clinical trial found that generative AI-powered chatbot therapy provided significant improvements in participants with depression and anxiety.
Those diagnosed with depression saw a 51% average reduction in symptoms, while participants with generalized anxiety registered an average reduction in symptoms of 31% - many of whom shifted from moderate to mild anxiety, or from mild anxiety to below the clinical diagnosis thresholds.

Doing therapy with my chatbot convinced me that AI can do extraordinary things.
The chatbot’s insights were shockingly accurate, almost as if it had pieced together parts of me that I hadn’t fully articulated myself.
I don’t think it’s a substitute for talking to human beings. The Dartmouth researchers don’t either, noting AI therapy “is still in critical need of clinician oversight.”
If you haven’t given it a shot, I’d recommend trying.

Brazilian fashion designer Lethicia Bronstein, known for handcrafted bridal wear, released a new collection using AI models on a virtual runway, writes Exame.
Designer Ricardo Almeida’s AI launch of new threads for Dia dos Namorados (Brazil’s equivalent of Valentine’s Day) had a decidedly Dali-esque presentation, also from Exame.

A widely-shared fake AI-generated video of showed Argentina’s former President Mauricio Macri announcing that the withdrawal of a candidate running in the Buenos Aires legislative election. A court ordered social media platform X to take down the video. (Crónica)
Brazil’s state of Goiás becomes the country’s first to approve AI regulations, putting itself ahead of the national legislation still being hashed out in Brazil’s chamber of deputies (Folha).
Brazilian energy firms are increasingly relying on AI to manage data around power consumption and generation, which is a critical part of the shift toward increasing the use of variable power such as solar energy while maintaining the stability of power grids. (Exame)
Paraguay’s Senate will begin discussing an bill governing the regulation of AI, with a focus on categorizing the risks of different AI systems (Última Hora).
The Abuelas de Plaza de Mayo are teaming up with AI researchers to help identify around 300 people who were taken as babies by Argentina’s 1976-1983 military dictatorship and put up for closed adoptions that cut them off from their biological relatives. (El País)

Artists call for AI Copyright Protections (BBC)
Over 400 UK artists including Elton John, Dua Lipa and Paul McCartney signed a letter urging Prime Minister Keir Starmer to maintain copyright laws to prevent AI companies from using their work without permission. Others say overly restrictive UK laws would slow AI development while doing nothing to prevent the use of such material in other countries.
Baidu Develops AI to Translate Animal Sounds (Reuters)
Chinese tech giant Baidu is working on an AI system designed to interpret animal vocalizations and behaviors, such as barks and meows, into human language.
U.S. Copyright Office Director Fired Amid AI Report Controversy (WaPo)
Shira Perlmutter, head of the U.S. Copyright Office, was dismissed shortly after releasing a report cautioning against the unchecked use of copyrighted material in AI training.