AI in Brazilian franchising: only 26% use it in a structured way, ABF finds

Artificial intelligence is already part of the conversation in Brazil’s franchise sector, but structured adoption remains limited. According to a 2025 survey by the Brazilian Franchising Association (ABF), cited in an article published by Pequenas Empresas & Grandes Negócios magazine on 27 September 2026, 26% of respondents use AI in a structured way. For 59%, the technology is either being tested or used at staff members’ own initiative.

From experimentation to organised use

The figures reveal two different stages of adoption. On one side is structured use, reported by roughly one in four survey participants. On the other is a larger group encompassing both experimentation and staff-led use.

This distinction matters when interpreting the findings. Testing a tool does not necessarily mean incorporating it into operations in an organised way. Similarly, a staff-led initiative may provide valuable experience, but the figure alone does not tell us whether it follows network-wide guidelines.

It is also important to recognise the limits of the available information. The published findings do not distinguish between respondents who are only testing AI and those whose staff use it on their own initiative. The 59% should therefore not be described either as an entirely informal group or as a set of projects coordinated by franchisors.

The survey’s main finding is the gap between experimenting with AI and using it in a structured way. The percentages alone do not establish which applications deliver returns or which networks are better prepared.

What the findings mean for discussions within franchise networks

For the franchise sector, the survey provides a starting point for discussing how to turn individual experiments into practices that can be evaluated. This is an implication of the findings, not a conclusion about the performance of the businesses surveyed.

A practical assessment can begin by mapping what is already in use. Which tools are being tested? What tasks do they support? Who monitors the results? These questions help distinguish between simply having access to a technology and integrating it into day-to-day work.

Another consideration is not to confuse frequency of use with quality. Using a tool every day does not, in itself, make it suitable for every task. Before expanding a trial, it is worth setting review criteria and making clear who is accountable for the final decision.

For franchisors and franchisees, the recommendation is to establish a shared vocabulary on the subject. Terms such as testing, approved application and structured use need clear meanings in discussions across the network. This groundwork makes it possible to discuss progress without treating every initiative as though it has reached the same stage.

AI agents on the horizon

The magazine article also includes an assessment by Cassio Pantaleoni, AI Director at Quality Digital, of the technology’s next developments. He identifies agent-based commerce, known as agentic commerce, and B2A infrastructure, described as a development enabling businesses to transact with AI agents, as among the most important trends.

This observation broadens the discussion beyond the current adoption of tools. By highlighting transactions with AI agents, Pantaleoni draws attention to potential changes in how commercial relationships are conducted.

It is important, however, to separate this outlook from the survey findings. Identifying a trend does not demonstrate that it has already been widely implemented across Brazilian franchise networks. The information presented does not quantify the adoption of agent-based commerce or B2A infrastructure within those networks.

The subject should therefore be treated as an area to monitor, rather than evidence of a transformation that is already complete.

Bringing the discussion into day-to-day operations

The contrast between the 26% reporting structured use and the 59% testing AI or relying on staff-led initiatives calls for a careful approach. Before adopting more complex solutions, each network can assess the stage its own initiatives have reached, without assuming that the survey percentages reflect its particular circumstances.

The practical route is to map existing uses, choose one application to evaluate, and assign responsibility and monitoring criteria. For anyone considering investing in a franchise in Brazil, it is worth asking how the network guides AI use. The priority need not be adopting more tools, but understanding what is already being done and how the results will be checked.