The 9th National Symposium for Mathematics Teacher Training, held from November 19 to 22, 2026, in São Luís, brought a critical message to the education sector: technology alone doesn't teach. Organized by the National Association of Mathematics Teachers in Basic Education (ANPMat), the event focused heavily on the necessity of capacitating teachers to properly integrate artificial intelligence into the classroom, particularly within mathematics instruction.

The Human Element in AI Adoption

Marcelo Viana, director-general of the Institute of Pure and Applied Mathematics (IMPA), emphasized that AI is no longer a distant concept but a rapidly advancing reality comparable to the telephone in its eventual indispensability. However, he warned that this technology is not neutral. Viana argued that the primary risk lies in not knowing how to use it, urging institutions to prepare educators so that AI contributes meaningfully to student formation rather than just automating existing processes.

Moving Beyond Plagiarism and Fraud

During public sessions, teachers raised concerns about AI-driven fraud, including the outsourcing of assignments and plagiarism across all educational levels. Victor Augusto Both Eyng, General Coordinator for Elementary Education at the Ministry of Education, addressed these fears by suggesting a shift in pedagogical strategy. He noted that students are already using these tools, so teachers must rethink traditional activities that yield identical, ready-made answers. Instead, Eyng proposed questions rooted in student identity, local context, and personal experiences to foster critical thinking.

Strategies for Critical Thinking

Eyng further outlined practical methods for students to develop authorship and critical analysis skills. He suggested that learners should verify information generated by AI, compare responses from different generative models, craft more qualified prompts, and rewrite content to create original productions based on the tool's output. The core message is to acknowledge AI's presence and work to qualify its use rather than ban it.

Sovereign AI and National Strategy

The discussion extended to Brazil's technological independence. Eyng referenced President Luiz Inácio Lula da Silva’s post-reelection speech in October 2026, which highlighted the need to invest in Math and Engineering to avoid dependency on foreign AI systems from China or the United States. As a concrete example, Viana cited SoberanIA, a Portuguese-language generative model launched by the Piauí state government in early 2026. Supported by the Ministry of Science, Technology and Innovation (MCTI) and the Ministry of Communications (MCom), this locally trained system aims to assist educators, automate services, and analyze large volumes of national data.

Key Takeaways

  • Teacher training is the bottleneck for effective AI integration in schools.
  • Pedagogical mediation is required to prevent AI use from becoming mere plagiarism.
  • Brazil is actively developing sovereign AI models like SoberanIA to reduce foreign dependency.
  • Educators are encouraged to design assignments based on student context rather than generic answers.

The Bottom Line

Without robust teacher training, AI in classrooms risks becoming a plagiarism engine rather than a learning tool. Brazil’s push for sovereign models like SoberanIA shows promise, but pedagogical strategy remains the true differentiator.