The return to classes drives demand for school supplies, electronics, and accessories between July and August. With the increase in sales, the opportunity for more sophisticated digital frauds also grows. According to the Brazilian Public Security Yearbook by DataSenado, in 2024, Brazilians suffered losses of over R$2.3 trillion due to cybercrimes. Visa Brazil reveals that the average ticket of fraudulent transactions was 60% higher than legitimate purchases in the same period.

In this scenario, Nethone, a digital fraud prevention solution, offers artificial intelligence models that analyze hundreds of user and device signals in each transaction with precision and without relying on manual review. The technology is trained with labeled historical data and continuous customer feedback, evolving and adapting to the specificities of each operation.

“Our focus is to empower the customer: the AI model does not replace the process but complements it. Innovation is done together, with constant learning based on real e‑commerce data,” explains Thiago Bertacchini, a fraud prevention specialist and Head of Sales at Nethone. He emphasizes that the effectiveness of AI is directly related to the quality and quantity of available data – the richer the transactional base, the more effective the defense against fraud.

The growing adoption of so-called agentic commerce — a practice in which AI agents perform automated tasks such as shopping and browsing — brings convenience but also unprecedented risks: exposure of credentials, injection attacks, and malicious bots that mimic human behavior. “With bots as well-trained as legitimate ones, fraud becomes harder to detect. Therefore, having a prevention approach that analyzes context and behavior, not just static rules, is essential,” warns Bertacchini.

One of the most critical challenges in back-to-school is to avoid false blocks to legitimate consumers, which can compromise conversion. Nethone uses proprietary technology to map thousands of signals in real-time — from behavioral patterns to variations in the purchase environment — and reduce false positives, maintaining a good flow in the shopping cart.

“Security doesn’t have to hinder sales. With AI and behavioral data, it’s possible to protect e-commerce without creating barriers to user experience — especially during high demand moments,” concludes the expert.