The newest teacher at San Diego charter school chain Altus Schools stands 6’2″, has bright blue eyes, and a bald head. It is also a robot.
Ameca, which the school touts as the “world’s most advanced AI-powered humanoid robot,” is the name of a pair of robots purchased by the school for a combined, eyebrow-raising figure of $500,000. The purchase is raising questions among parents and community members.
Altus expects the ChatGPT-enabled robots to be onsite this fall. Principal Cathryn Rambo wrote in an email to families that she was “thrilled to be the first school in the world researching the use of physical AI as a teaching partner,” according to Voice of San Diego, which reviewed the email.
Altus charter schools are designed for students who have fallen behind academically. They are given the chance to fast-track classes and catch up. The schools, particularly the San Diego location, have been credited with helping students succeed at impressive rates.
Students typically work independently, but they can go to one of several resource centers for one-on-one lessons. That is where the robots come in.
Ameca can switch between four different personas for students. Three of them, Sage the Teacher, Ari the College and Career Planner, and Lexi the Translator, are about what you would expect from a school-based robot. The fourth, though, a persona called “Remi the Wellness Coach,” is raising concerns.
Students who struggle academically are often from at-risk environments, with higher stress and anxiety levels and possible self-esteem issues. Some might struggle socially. That can make them more susceptible to forming an unhealthy or harmful relationship with AI.
A 2025 study by Common Sense Media and Stanford University found that leading AI companion platforms pose “very serious risks” for teens by simulating real relationships and creating emotional bonds. That can worsen mental health struggles and discourage real-world friendships.
“Harmful content is common, even with safeguards,” the report reads. “Our testing found that many platforms still allow inappropriate conversations, unsafe advice and unhealthy emotional reinforcement—even in ‘teen mode,'”
Altus’ Rambo told Voice of San Diego that Ameca will not replace traditional mental health services. She said the wellness coach persona would instead offer encouragement to students who are anxious about upcoming tests and similar scenarios.
“If a student is upset about an argument with a parent, we’re never going to put them in front of a robot,” Rambo said.
Additionally, children will not be left alone with the machines. And to protect the privacy of students, the robots’ memory is erased after every interaction. No data is recorded, either.
The robots are programmed to avoid certain topics. For example, Ameca can discuss the Clinton presidency, but if asked about the sex scandal that erupted in 1998, it is programmed to be less specific. It can also imitate certain people, but controversial figures are off limits.
AI isn’t perfect, though. And while new models are less prone to hallucination and stay on topic better than earlier ones, they can still be tricked into speaking beyond their programmed boundaries at times.
Beyond the use of AI robots in a school for students who have fallen behind, there are other questions. If the school is researching the use of AI as a teaching partner, as Rambo said in the email, who is it working with on the study? And was it necessary to spend $500,000 at a time when many schools are underfunded?
There is also the question of why the school needed a humanoid robot if the real focus is on AI as a teaching partner. A subscription to ChatGPT, or even a curated version of that chatbot that purports to filter out inappropriate content, would cost significantly less.
Those are questions the school is likely to face at the next round of parent-teacher conferences.