巴西七成网民向AI交隐私，66%忧滥用，中资企业AI治理需设红线

Brazilians’ use of artificial intelligence is not limited to professional and academic matters. Although 73% use AI platforms for work and study, half or more share personal preferences, feelings, health information, and photos of themselves or people they know with these services. The figures are part of the first mapping of AI use by Brazilians, conducted as part of the third edition of the “Privacy and Personal Data Protection Survey” by the Regional Center for Studies on the Development of the Information Society (Cetic.br), with support from the National Data Protection Authority (ANPD). The survey conducted 2,500 online interviews with internet users aged 16 and older. In addition to work and study, 58% of respondents share “personal preferences and tastes, such as movies or music” with AI chatbots, 54% discuss “feelings and emotions,” 52% address “health information, such as symptoms and test results,” and 50% share “photos of themselves or people they know” during interactions. The least-cited categories of AI use were “personal finances, such as budgets and investments” (41%) and “personal information, such as name, address, date of birth, or tax ID” (37%), according to the survey conducted in December 2025. That behavior contrasts with the concern expressed by 66% of respondents about how the companies behind AI platforms use their information. Concern was higher among those who use AI on both mobile phones and computers (68%) and respondents with higher levels of education (72%). The findings warn companies that they need governance over AI applications in the workplace. “These interactions may involve the sharing of organizations’ strategic information, making it increasingly necessary for companies to advise employees about the limits on what information should be entered into these platforms,” Leonardo Lins, one of the coordinators of the Cetic.br study, told Valor. Important personal-data management practices among consumers declined compared with previous studies. In the current survey, 65% said they had read a website’s or application’s privacy policy, compared with 67% in 2023 and 68% in 2021. The share of respondents who checked whether a website displayed a security padlock fell from 70% in 2021 to 67% in 2023 and 63% in the survey’s third edition. The study also presents troubling data on online scams. Among respondents, 32% had experienced attempted scams using their personal information, equivalent to 45 million Brazilians, while 20% said they had fallen victim to this type of crime. Messaging applications were the main channel used to contact potential victims, cited by 84% of respondents, followed by telephone calls (79%), fraudulent websites or applications (71%), text messages (71%), email (69%), and social media (67%). In education, the survey found that 39% of public-school administrators cited privacy and personal data protection concerns as a main challenge to adopting school-management software. Among respondents, 25% said data security was one obstacle to using devices, platforms, or other digital educational resources. Cetic.br is the division of the Brazilian Network Information Center (NIC.br) responsible for producing indicators and statistics on the availability and use of information and communications technologies in Brazil. NIC.br is linked to the Brazilian Internet Steering Committee (CGI.br). Translation: Todd Harkin