Blab AI
Brazils Hospitals Reach 18% AI Adoption, Backed by New Medical Council Rules
In 2025, 18 % of Brazil’s health‑care facilities reported using artificial intelligence (AI), a steep rise from 4 % the year before, according to the TIC Saúde 2025 survey. The jump reflects growing use of generative language models, text mining, workflow automation and speech‑recognition tools in hospitals across the country.
The survey, released in May 2026, shows that AI adoption is strongest in larger hospitals with more than 50 inpatient beds and in units that provide Diagnostic and Therapeutic Support Services (SADT) such as radiology and ultrasound. These environments generate higher volumes of clinical data and possess more advanced technology infrastructure, making them better suited for AI integration.
The Federal Council of Medicine (CFM) issued a rule in February that regulates doctors’ use of AI. The resolution, which will take effect 180 days after publication, allows clinicians to use AI as a support resource but requires that the final clinical decision remain theirs. It also mandates that patients be informed whenever AI is employed and that doctors cannot be held unduly responsible for AI failures, provided the technology is used critically and ethically. The rule will compel health‑care institutions to establish an AI and telemedicine committee.
A bill that would establish a national regulatory framework for AI has passed the Senate and is now under discussion in the Chamber of Deputies. The draft legislation classifies AI systems that pose a significant risk to physical or mental integrity as “high risk” and requires companies, professionals and sectors to develop their own rules for use. The bill also permits the creation of nonprofit entities to set standards and share best practices, subject to competition law. The bill’s rapporteur, Congressman Aguinaldo Ribeiro, has indicated that he intends to remove the risk‑classification section from the text.
In practice, AI is already being used to streamline hospital operations. CTC Tech’s Lya Health listens to and transcribes conversations between patients and doctors, organizes medical records, and suggests questions and possible diagnoses. According to Valter Lima, CEO of CTC Tech, the system does not store conversation data on the company’s platform; it is deleted after the consultation and is not used for learning. Lima said the technology allows doctors to focus more on patients while improving record quality.
Munai’s platform integrates with electronic health records and generates alerts for patients who may need to be discharged or whose condition may worsen. Infectious‑disease specialist Hugo Morales, co‑founder and medical director of Munai, explained that the system does not provide a diagnosis but indicates a patient’s trajectory based on 120,000 analyzed cases and the hospital’s protocol. The alerts aim to reduce serious events such as cardiorespiratory arrests.
The federal government has announced the creation of a National Network of Smart Hospitals and Services of the Unified Health System (SUS). The network will invest in AI, big‑data analytics and other digital technologies, with a focus on automating patient triage and optimizing appointment scheduling in a future smart hospital in São Paulo.
Experts note that acceptance by clinicians and patients remains the first hurdle to broader AI deployment. Professor Alexandro Chiavegatto of the University of São Paulo’s School of Public Health says that digital systems are increasingly used to answer questions and retrieve information, and that the SUS already leads in data availability compared with private hospitals, which invest more heavily in technology.
A persistent challenge is the shortage of qualified human capital needed to train algorithms and expand national databases. Chiavegatto and data‑privacy specialist Rafael Zanatta both emphasize that building robust AI models requires skilled professionals, a profile also in demand in the financial sector.
Overall, Brazil’s AI adoption in health care is still in early stages, but the combination of regulatory guidance, growing institutional use, and emerging commercial solutions suggests a trajectory toward wider integration. The next few months will see the CFM resolution take effect, the AI bill move through the Chamber, and further deployments of AI‑enabled triage and record‑keeping systems in public and private hospitals.
The survey, released in May 2026, shows that AI adoption is strongest in larger hospitals with more than 50 inpatient beds and in units that provide Diagnostic and Therapeutic Support Services (SADT) such as radiology and ultrasound. These environments generate higher volumes of clinical data and possess more advanced technology infrastructure, making them better suited for AI integration.
The Federal Council of Medicine (CFM) issued a rule in February that regulates doctors’ use of AI. The resolution, which will take effect 180 days after publication, allows clinicians to use AI as a support resource but requires that the final clinical decision remain theirs. It also mandates that patients be informed whenever AI is employed and that doctors cannot be held unduly responsible for AI failures, provided the technology is used critically and ethically. The rule will compel health‑care institutions to establish an AI and telemedicine committee.
A bill that would establish a national regulatory framework for AI has passed the Senate and is now under discussion in the Chamber of Deputies. The draft legislation classifies AI systems that pose a significant risk to physical or mental integrity as “high risk” and requires companies, professionals and sectors to develop their own rules for use. The bill also permits the creation of nonprofit entities to set standards and share best practices, subject to competition law. The bill’s rapporteur, Congressman Aguinaldo Ribeiro, has indicated that he intends to remove the risk‑classification section from the text.
In practice, AI is already being used to streamline hospital operations. CTC Tech’s Lya Health listens to and transcribes conversations between patients and doctors, organizes medical records, and suggests questions and possible diagnoses. According to Valter Lima, CEO of CTC Tech, the system does not store conversation data on the company’s platform; it is deleted after the consultation and is not used for learning. Lima said the technology allows doctors to focus more on patients while improving record quality.
Munai’s platform integrates with electronic health records and generates alerts for patients who may need to be discharged or whose condition may worsen. Infectious‑disease specialist Hugo Morales, co‑founder and medical director of Munai, explained that the system does not provide a diagnosis but indicates a patient’s trajectory based on 120,000 analyzed cases and the hospital’s protocol. The alerts aim to reduce serious events such as cardiorespiratory arrests.
The federal government has announced the creation of a National Network of Smart Hospitals and Services of the Unified Health System (SUS). The network will invest in AI, big‑data analytics and other digital technologies, with a focus on automating patient triage and optimizing appointment scheduling in a future smart hospital in São Paulo.
Experts note that acceptance by clinicians and patients remains the first hurdle to broader AI deployment. Professor Alexandro Chiavegatto of the University of São Paulo’s School of Public Health says that digital systems are increasingly used to answer questions and retrieve information, and that the SUS already leads in data availability compared with private hospitals, which invest more heavily in technology.
A persistent challenge is the shortage of qualified human capital needed to train algorithms and expand national databases. Chiavegatto and data‑privacy specialist Rafael Zanatta both emphasize that building robust AI models requires skilled professionals, a profile also in demand in the financial sector.
Overall, Brazil’s AI adoption in health care is still in early stages, but the combination of regulatory guidance, growing institutional use, and emerging commercial solutions suggests a trajectory toward wider integration. The next few months will see the CFM resolution take effect, the AI bill move through the Chamber, and further deployments of AI‑enabled triage and record‑keeping systems in public and private hospitals.
