New Twilio report highlights behaviors and challenges involving conversational AI for the Brazilian and Latin American market

Twilio, a customer engagement platform that drives personalized real-time experiences for today's top brands, releases the 1st edition of its “ report“Inside the Conversational AI Revolution”¹. The document highlights the main trends, behaviors and challenges related to adoption, perceptions and expectations about conversational AI, and presents important insights for Latin America and Brazil.

In its first edition, the report positions Latin America as the world leader in the implementation of AI for customer service. 31% of companies in the region claim to have already completed the development and full implementation of conversational AI for customer service, exceeding the global average of implementation, which is currently 28%. In addition, Brazil stands out in the region, with 44% of local companies claiming to have already completed or are in the final phase of implementation and development of conversational AI for customer service.

“Globally, the main use cases of AI for CX are simple queries, such as obtaining product information, checking order status, and communicating with customer service representatives through messaging apps, especially WhatsApp”, says Jose Eduardo Ferreira, Regional Vice President Sales for Latin America. “Latin American consumers already interact with AI through messaging apps three times more than the global average, which represents a huge opportunity for Brazilian companies to utilize this channel to better connect with their customers.”

Other important data presented in the report include:

• Latin America has created its own pace of innovation, consolidated by the real impact these solutions have on the customer: most organizations already consider switching AI solutions in less than 12 months.There is no attachment to technologies that become obsolete quickly due to the creation of new hardware and more advanced software;
• Companies refuse to use a single AI model: 94% of Latin American companies follow a “multimodel” approach, whether due to solution categories, use cases or a combination aiming at better performance in a hybrid model;
• Challenges persist in relation to customer relationships: there is a paradox when discussing company perception and real customer experience with AI. Companies believe they offer an effective automated service, but consumers report flaws, especially in the continuity of the conversation, particularly in the transition between the automated agent and the human agent. While 92% of organizations in Latin America claim that their customers are somewhat or very satisfied with their conversational AI, only 61% agree. In Brazil, the disparity is from 96% to 66%. Companies should work to close this gap.

Consumers value effective and fast resolutions above all else, and their willingness to interact with AI agents is highly conditioned on the performance of the AI agent.

Companies looking to bridge the customer satisfaction gap with AI and deliver trusted, empathetic, and personalized experiences should focus on: effective problem solving, not just speed; improved transitions between AI agents and human agents; enhanced security, privacy, and transparency to build customer trust; and adaptive omnichannel solutions with a flexible, composable technology infrastructure that supports multiple AI models.

As these companies prepare to replace their current solutions in the next 12 months, additional investments in interoperability and modularity will play a key role in strengthening their adaptability in the more distant future.

The full report can be viewed here.

Methodology

Twilio surveyed 4,800 global consumers and 457 business leaders (full-time directors or senior positions at B2B and B2C companies) in 12 countries between August 7 and September 4, 2025, with 3 more countries surveyed between October 10 and 17, 2025.

¹A Twilio conducted this research using an online questionnaire prepared by Method Research and distributed by RepData. Data were collected from August 7 to September 4, 2025, with an additional survey in 3 countries conducted from October 10 to 17, 2025.