FTD to invest R$100m a year in educational AI, without ‘big techs’
Sandro Bonás
Gabriel Reis/Valor
FTD, the Marista Group’s educational publisher specializing in textbooks and learning systems, will more than triple its annual technology investment from the current R$30 million to R$100 million beginning in 2027. The additional funding is intended to “control” the use of artificial intelligence in education and prevent distorted educational content from reaching students.
Act Digital to invest R$100m to expand AI deployment in Brazil
AI helps Elenco Digital recruit actors
World Cup offside system gets AI upgrade
“When AI generates text, we don’t know what values and interests are behind that content. The entire world is built on history, which we tell according to editorial standards. Now these new synthetic stories are being told according to the interests not of education companies, but of technology groups. OpenAI, Google, and Anthropic are defining the narrative. Elon Musk, through his company X, is producing content,” said Sandro Bonás, FTD’s new chief executive, who took office in June.
Bonás warned that publishers need to maintain control over educational content to avoid the risk of history being distorted. Last week, Elon Musk wrote on X (formerly Twitter) that “before the end of this year, Grok Imagine [his AI] will produce a historically accurate feature-length adaptation of The Odyssey faithful to Homer’s work.” The post followed the release of a film that drew criticism for taking creative liberties.
According to FTD’s chief executive, one of the key paths forward for the industry is for publishers to transform themselves into educational technology companies with proprietary AI. He said such systems should foster students’ cognitive development through a Socratic methodology that encourages questioning and critical thinking. Current AI tools, he argued, simply provide ready-made answers, limiting the development of the mental processes essential to learning. Today, an estimated 84% of students use AI tools to complete school assignments.
Education specialists have expressed concern about the cognitive development of today’s students, who already experienced a learning gap during the Covid-19 pandemic. They warn that two consecutive periods of educational setbacks could have lasting consequences for an entire generation.
According to Bonás, the sharp increase in investment reflects both urgency and concern over “preventing technology groups from entering a field that we need to preserve, with its values and ethical standards. We want to lead this process given our scale.”
The Marista publisher accounts for 25% of the textbooks distributed through Brazil’s public education system and serves 7,000 private schools across the country. It prints 150 million copies annually. In 2025, FTD posted revenue of R$1.5 billion and aims to double that figure by 2030.
In addition to the publishing business, the Marista Group owns the Pontifical Catholic University of Paraná (PUC-PR), 65 private schools under the Marista and Champagnat brands, and another 34 tuition-free schools. In healthcare, the group also operates the private São Marcelino Champagnat Hospital and Cajuru University Hospital, affiliated with PUC-PR, both located in Curitiba. The group’s total revenue reached R$3.3 billion in 2025.
Bonás, who was recruited for the role because of his expertise in AI, believes revenue growth will be driven by new educational materials incorporating the technology.
The company also plans to use AI to personalize learning according to each student’s needs, such as providing more targeted support in subjects where they face greater difficulty or preparation for specific university entrance exams.
The first step will be AI literacy programs for teachers, students, and families. Bonás advocates making AI literacy a mandatory subject from the first year of elementary school, when children are six years old.
UNESCO is among the organizations promoting digital literacy, which goes beyond technical knowledge to include the ability to think critically about technology.
“Children need to understand that when they’re on TikTok, Netflix, or YouTube, they’re receiving recommendations generated by AI algorithms. Today, more than 80% of what people watch on YouTube is not something they actively chose. In other words, people have lost the autonomy to decide what they want to watch. That’s crazy,” Bonás said.
He also expressed concern about Brazil’s dependence on technologies developed abroad. In AI, the market is concentrated among companies from the United States and China. He noted that there is growing global debate over the lack of technological sovereignty, the risks of potential disruptions in access, and the exposure of confidential data.
“In Europe, they introduced regulation to reduce this type of dependence and encourage the development of European companies,” he said. He added that excessive regulation could hinder innovation, while an overly permissive environment, such as in the United States, could create risks by allowing still-immature technologies to advance unchecked.
Contacted by Valor, Google declined to comment. OpenAI, X, and Anthropic did not respond to requests for comment.
