AI in healthcare: uses, examples and rules in the Philippines
Short answer
The best-documented use of AI in Philippine healthcare is chest X-ray reading for tuberculosis screening, with a DOH and WHO statement from November 2025 and two published studies. The Data Privacy Act treats health data as sensitive personal information, and the FDA has a draft circular on medical device software. No health-specific AI adoption rate was found. This page informs and does not give medical, legal or purchasing advice.
On this page
Key facts
- Best-documented Philippine use
- AI-assisted chest X-ray reading for tuberculosis screening
- DOH and WHO statement on AI-powered portable X-rays
- 12 November 2025
- Number needed to screen, AI pathway vs human readers (2024 paper)
- 26.3 vs 41.5
- Health data under the Data Privacy Act
- Sensitive personal information, Section 3(l)
- FDA Philippines on medical device software
- Draft circular, comments to 20 July 2025
- Health-specific AI adoption rate
- Not found from a primary source
This page reports what published sources say about AI in Philippine healthcare. It does not advise on diagnosis, treatment, product choice or legal compliance (see “What this page does not advise on”). Every figure carries its source and date. All sources were read on 12 Oct 2026.
Adoption context
The most recent firm-level survey this site found covers all sectors, not health. The 2021 PIDS Survey of Innovation Activities of Establishments, analysed in PIDS Discussion Paper 2024-35 (December 2024), puts AI adoption at 3.02 percent of establishments overall. The sector breakdown reproduced on this site covers BPO, ICT and agriculture, not health. The report on AI in the Philippines sets out that survey, a second PIDS figure of 14.9 percent that the paper does not reconcile with it, and the sources behind both.
For health, what is published is policy direction, not adoption counts.
- On 26 February 2026, DOH spokesperson Undersecretary Albert Domingo said AI is part of the DOH digital health strategy and needs reliable, interconnected health data systems. He named local health information exchanges as the priority, starting with Aklan, then Batangas, Benguet, Laguna and Sarangani (GMA News, 26 February 2026, used only to locate the statement).
- At a DOH workshop on 11 and 12 December 2023, the director of the DOH Knowledge Management and Information Technology Service said the DOH was committed to developing a Digital Health Strategic Framework 2024 to 2028 and a Health Sector Enterprise Architecture 2023 to 2035 (University of the Philippines Manila, 2 January 2024).
- On 1 June 2026, the National Academy of Science and Technology, Philippines (NAST PHL) held a focus group on AI-driven healthcare. Participants noted gaps in the law on digital health and AI governance, urged a national health information exchange framework, and recommended that the FDA evaluate AI-enabled medical devices before deployment (NAST PHL, published 14 July 2026). These are participants’ recommendations, not rules.
Administrative uses
Administrative automation covers work around care, not care decisions. The WHO guidance on large multi-modal models (18 January 2024) lists clerical and administrative work as one of five broad health applications, for example documenting and summarizing patient visits in electronic health records. It also names risks: false or incomplete statements, automation bias (errors that go unnoticed) and cybersecurity weaknesses that could expose patient information.
No Philippine study of the four tasks below was found. They are described in general terms, and each description says what the software does, not what it achieves.
Appointment handling. Software takes booking requests, offers open slots and confirms or reschedules a visit. An appointment record carries a name, contact details and often the reason for the visit. A reason for a visit can be health information, which the Data Privacy Act treats as sensitive. Staff should know what the system stores and where.
Patient messaging. A chatbot or message workflow sends reminders, answers opening-hours questions and routes a message to the right person. Messages that describe symptoms are health information. Because language models can state false things, a clinic decides in advance what the software may say and who reads the cases it cannot handle. The guide to AI automation explains how a human-in-the-loop step fits into a workflow.
Claims and records. Section 36 of the Universal Health Care Act (Republic Act 11223, approved 20 February 2019) requires health service providers and insurers to maintain a health information system with electronic health records and an electronic prescription log consistent with DOH standards, and to uphold patient privacy under the Data Privacy Act. Software that drafts or checks entries sits inside that duty. The Act does not mention AI.
Scheduling. Staff rosters and room bookings are operations data, lower in privacy risk unless they identify patients. No Philippine source read describes AI scheduling in a health setting.
Clinical-support uses
Clinical-support tools assist a licensed professional. This page does not describe them as replacing one.
Imaging. The published Philippine evidence is here, in tuberculosis (TB) screening. Software reads a chest X-ray and flags findings that suggest TB, and a confirmatory laboratory test follows. The next section gives the sources and their limits.
Triage support. DOH-NPC Joint Memorandum Circular 2020-0001 lists triaging among the services a DOH telemedicine partner can provide. It does not mention AI. No Philippine source read describes AI triage at a named provider.
Documentation. The WHO guidance gives visit summaries as an example. No Philippine source read reports a deployment. A clinician reviews any draft, because the WHO names automation bias as a risk.
Philippine examples
Each example comes from an organization’s own page or a published paper, with its date.
| Example | What the source says | Date and source type |
|---|---|---|
| DOH TB screening with AI-powered portable chest X-rays | Health Secretary Teodoro Herbosa said the country is already using ultra-portable AI-powered chest X-rays alongside WHO-recommended molecular tests. In the Cordillera region, health workers bring AI-powered portable X-ray units closer to communities. The release sets a target of screening 12 million Filipinos by 2026. | 12 November 2025, joint WHO and DOH release on the WHO Philippines site |
| ACCESSTB case finding, Philippine Business for Social Progress (PBSP) | A retrospective analysis of PBSP reports from three regions compared an AI pathway with a human-reader pathway. Number needed to screen: 26.3 with AI, 41.5 with human readers. Dropout after the X-ray: 16.6 and 43.1 percent. Positive individuals who received treatment: 3.8 and 2.4 percent. The authors link the lower dropout to faster X-ray results and say cost-effectiveness needs more research. | February 2024, International Journal of Health Sciences and Research |
| qXR 3.0 in two Metro Manila primary care clinics | A prospective accuracy study in Marikina and Mandaluyong (October 2021 to February 2022) enrolled 82 adults. Against a molecular test, sensitivity was 100 percent (95 percent confidence interval 47.8 to 100) and specificity 72.7 percent (49.8 to 89.3). The authors say it meets the WHO specificity threshold of 0.70. | 31 January 2025, Acta Medica Philippina |
| DOST Health Technology Assessment of Huawei Cloud AI-Assisted Diagnosis for COVID-19 | The page asks whether the DOH should join a proof-of-concept study. It states no conclusion, recommendation or decision. | 13 April 2020, DOST HTA page |
Swipe the table sideways to see all columns.
Limits, as the authors state them. The PBSP analysis was not randomized, used different barangays for each approach and had no radiologist review of films the software read as negative. One author is affiliated with Qure.ai, which makes the software, and the authors declare no conflict of interest. The Metro Manila study fell short of its planned sample size, so its estimates are imprecise, and the authors say the results may not generalize. Qure.ai also hosts a summary of the PBSP paper. It is a company page and supplies no figure here.
Hospital and startup announcements were searched for. The ones found were vendor or news coverage with no primary page from the provider, so none is listed.
Rules that apply
This section summarizes documents by number and date. It is not legal advice.
Data Privacy Act and sensitive personal information. Republic Act 10173, approved 15 August 2012, defines sensitive personal information in Section 3(l). The definition includes information about an individual’s health and records issued by government agencies such as previous or current health records. The National Privacy Commission issued NPC Advisory No. 2024-04 on 19 December 2024, on the application of the Act to AI systems that process personal data. The NPC site blocks automated access, so the advisory is reported by number, date and subject only.
DOH and FDA positions on software and devices. FDA Philippines has posted a draft circular, Guidelines on the Regulation of Medical Device Software, with a comment deadline of 20 July 2025. It supplements DOH Administrative Order 2018-0002 and aims to help decide whether software is a medical device, assign a risk class and set application requirements. It includes a category for artificial intelligence medical devices. Risk class follows intended use: Class A devices need notification, and Classes B, C and D need registration. Whether the circular has been finalized was not confirmed.
Telemedicine guidance. DOH-NPC Joint Memorandum Circular 2020-0001 (28 March 2020) set telemedicine guidelines for the COVID-19 response. Its principles include that face-to-face consultation remains the gold standard for clinical care, that emergency and serious conditions should not be managed through telemedicine, and that providers apply the minimum security standards of the NPC and the DICT. It requires informed consent covering privacy concerns and the risk of technology failure. DOH, DILG and PhilHealth Joint Administrative Order 2021-0001 (19 May 2021) then made telemedicine a mode of delivering individual-based health services under the Universal Health Care Act. Whether either has been amended was not confirmed.
Ethics guidance. The WHO report Ethics and governance of artificial intelligence for health (28 June 2021) sets six principles: protecting human autonomy; promoting well-being, safety and the public interest; ensuring transparency, explainability and intelligibility; fostering responsibility and accountability; ensuring inclusiveness and equity; and promoting AI that is responsive and sustainable. It is guidance, not Philippine law.
Opportunity list
These are options for a provider to evaluate. Each is tied to a source or labelled an illustration. None is a recommendation.
| Provider size | Option to evaluate | Basis |
|---|---|---|
| Clinic | Reminder and booking messages, with a rule on what the software may say and a person to take over | Illustration. Privacy duties: Data Privacy Act, NPC Advisory No. 2024-04 |
| Clinic | Telemedicine through a platform that meets consent and security principles | JMC 2020-0001; JAO 2021-0001 |
| Hospital | Drafting visit summaries from clinician input, with clinician review | WHO guidance, 18 January 2024. No Philippine deployment found |
| Hospital | Software that flags chest X-rays for TB follow-up, after checking its FDA status | Flores 2025; Nampewo 2024; FDA draft circular |
| Network | Shared records and exchange between facilities, before any AI layer | Republic Act 11223, Section 36; DOH statement of 26 February 2026 |
| Network | Pilots with local validation and checks on results | Illustration. NAST PHL focus group urged continuous regulation and human oversight |
Swipe the table sideways to see all columns.
What changes for roles and tasks
Few sources describe role changes, so this section is short.
- Physicians. JMC 2020-0001 says the physician who first sees a patient referred to a facility stays responsible for coordinating care. The NAST PHL focus group held that AI must be complemented by human judgment.
- Administrators. Section 36 of Republic Act 11223 gives providers a duty to maintain health information systems. NAST PHL participants assigned health data governance to the DOH and PhilHealth.
- Front desk and nurses. No source read describes how AI changes these roles in the Philippines, so the page does not.
What this page does not advise on
This page does not advise on diagnosis, treatment or any clinical decision, and it does not replace a licensed professional. It does not advise on which product or vendor to choose. It does not advise on legal compliance: the rules are summarized from documents read on 12 Oct 2026, and a provider should ask a qualified lawyer or its data protection officer about its own case. No health professional, lawyer or other qualified person has reviewed this page, so it carries no reviewer line.
What is not published
Each item was true of the sources reviewed on 12 Oct 2026.
- A primary-source count of Philippine hospitals, clinics or practitioners using AI.
- Any Philippine study of AI for appointment handling, messaging, claims, scheduling, triage or documentation.
- A hospital or startup page announcing a dated AI deployment.
- A DOH-hosted copy of the DOH statement on AI and health information exchanges, and the Digital Health Strategic Framework 2024 to 2028 as a published document.
- The outcome of the DOST HTA review, and the final FDA circular on medical device software.
- An Ateneo de Manila cost model of AI-assisted TB screening reported in news in September 2026. The paper could not be read, so no figure from it is used.
- The text of NPC Advisory No. 2024-04 (not opened by this site’s tools) and of DOH-NPC JMC 2020-0002 (not opened).
Method
Each example is attributed to the organization’s own page or a published paper. News reports only locate a statement and are marked. Figures are shown as printed with their period. The two scanned DOH circulars were read as page images. Studies authored or funded by a product’s maker are flagged. The page reports what sources state and adds no clinical interpretation. No qualified health or legal review has taken place. Corrections follow the editorial policy.
Frequently asked questions
How is AI used in Philippine hospitals and clinics?
The best-documented use is reading chest X-rays for TB screening. The DOH and WHO said on 12 November 2025 that AI-powered portable X-rays are in use, and two published studies examine it. Published Philippine sources on other uses were not found.
Which rules govern AI in Philippine healthcare?
No law written for AI in healthcare was found. The Data Privacy Act, NPC Advisory No. 2024-04, DOH telemedicine issuances, the Universal Health Care Act and FDA device rules may apply, depending on the use. This page does not give legal advice.
How is patient data protected when AI is used?
The Data Privacy Act treats health information as sensitive personal information, and Section 36 of the Universal Health Care Act requires health information systems to uphold that privacy. How a provider meets these duties is a question for its data protection officer.
Which healthcare tasks suit AI first?
The sources reviewed do not rank tasks. Philippine evidence is concentrated in TB X-ray screening, and the WHO lists clerical work among five broad applications. A provider decides by testing a narrow task with human review.
Sources
All sources were read on 12 Oct 2026.
- Republic Act No. 10173, Data Privacy Act of 2012, approved 15 August 2012.
- Republic Act No. 11223, Universal Health Care Act, approved 20 February 2019.
- National Privacy Commission, NPC Advisory No. 2024-04, 19 December 2024.
- DOH and NPC, Joint Memorandum Circular 2020-0001, Guidelines on the Use of Telemedicine in COVID-19 Response, 28 March 2020 (scanned copy).
- DOH, DILG and PhilHealth, Joint Administrative Order 2021-0001, Guidelines on the Implementation of Telemedicine in the Delivery of Individual-Based Health Services, 19 May 2021 (scanned copy).
- Food and Drug Administration Philippines, Draft for Comments: Guidelines on the Regulation of Medical Device Software, comments to 20 July 2025.
- World Health Organization, Ethics and governance of artificial intelligence for health, and news release with the six principles, 28 June 2021.
- World Health Organization, AI ethics and governance guidance for large multi-modal models, 18 January 2024, and the guidance page.
- WHO Philippines and DOH, WHO, DOH target 12M Filipinos to be screened for tuberculosis by 2026, 12 November 2025.
- Flores GP and others, Diagnostic Performance of a Computer-aided System for Tuberculosis Screening in Two Philippine Cities, Acta Medica Philippina 59(2), 31 January 2025.
- Nampewo I, Ariana P, Vijayan S and Holthof B, Benefits of Artificial Intelligence versus Human-Reader in Chest X-ray Screening for Tuberculosis in the Philippines, International Journal of Health Sciences and Research 14(2), February 2024.
- National Academy of Science and Technology, Philippines, NAST PHL leads dialogue on advancing AI-driven healthcare, focus group of 1 June 2026, published 14 July 2026.
- DOST Health Technology Assessment, Huawei Cloud AI-Assisted Diagnosis, 13 April 2020.
- University of the Philippines Manila, workshop report on the national eHealth program, 2 January 2024.
- GMA News Online, DOH on AI tools in healthcare services, 26 February 2026 (news report).
- Philippine Institute for Development Studies, Discussion Paper 2024-35, December 2024.