The World Health Organization has appealed to governments to make available national strategies dedicated to artificial intelligence in health care and more robust legal structures, warning that the rapid introduction of the technology has exceeded the capacity of measures that are in place to guard the interests of the patients and the health workers. In an extensive report on the state of readiness for AI in the WHO European Region, the officials pointed out that the countries were lagging for governance, i.e. making and enforcing laws, regulations and policies related to AI that are health-oriented.
Only four of the 50 Member States (8 percent) in total that replied are currently preparing AI national strategies which focus on health, and a further seven (14 percent) are about it, though not yet finalized. In almost every country (86 percent), legal uncertainty was listed as the main reason preventing the introduction of AI technologies and less than 10 percent of the countries have liability regulations that specify who, how and when, an AI system is liable to be blamed if it harms or makes mistakes.
“We find ourselves at a crossroads, ” said Dr Natasha Azzopardi-Muscat, Chief of the Health Governance Unit at WHO/Europe. “AI can be used for enhancing people’s health and well-being, lightening the workload of our overworked health workers and cutting down healthcare expenditures, or conversely it could be a tool that harms, destroys patients’ privacy and leads to care inequities, The choice is ours to make.”
Mentioning this AI readiness report, based on findings obtained in the period June 2024 – March 2025, WHO/Europe pointed out that AI is already serving multiple purposes in different contexts, for example diagnostics administrative jobs, tracking diseases and communication with patients within the region Still investment, education and regulatory clarity are still not there. Nearly 8 out of 10 countries stated that money was a significant limiting factor. And fewer than one-fourth offer regular courses on AI for health workers.
Possible use of AI in medical decisions by clinicians would not only create the question of responsibility but ethical as well, because of the limitations on understanding that systems can have. This situation is leading to a reluctance in the more widespread use of AI by medical practitioners. WHO/Europe urges countries to carry out concrete actions. Some among these are developing health-specific AI strategies that correspond to public health objectives, establishing liability system and setting up procedures in place so patients who suffered adverse events after using AI can be reimbursed or other forms of redress made available.
Also, the organization advocates the use of systems only after thorough testing about their effectiveness, absence of bias, and safety. And, the need for investment in the development of a workforce skilled in AI is emphasized with open and public engagement, better control of data by people and between people (which includes international arrangements), and ethical rules to ensure that a bias or infringement of privacy is avoided.
Dr.Hans Henri P.Kluge, WHO Regional Director for Europe, stated that, if these conditions are not met, the new technology will likely end up exacerbating health disparities rather than mitigating them. “We could be the cause of making health disparities worse if a clear roadmap for AI is lacking as well as the legal safeguards on the issue of privacy, the protection of personal information, and the development of people’s knowledge about AI, ” he remarked. The new results support the previous WHO policy documents, like the 2021 ethical principles on AI, or the 2023 regulatory considerations paper focusing on transparency, risk mitigation, data quality, privacy protection and multi-stakeholder cooperation.
The officials added that the laws on AI, on the general level, cannot be enough to cover the issues of health data sensitivity besides those of clinical diagnosis and prescription. In the health sector, specific rules shall tackle, among others, the continuous learning of an AI through the use, the liability of each of the parties responsible for the design, delivery, or use of the system, and the possibility that biased data during training will lead to an uneven distribution of outcomes to various segments of the population who will receive the AI-generated care.
In addition to its spotlight on the European Region, the findings of the report will have impact worldwide through AI tools in diagnostics, drug development as well as health system management that are spreading at lightning speed. WHO keeps giving technical advice, building capacities, and providing international regulatory dialogue platforms as it wants the use of AI not to be detrimental but beneficial to the delivery of universally accessible and high-quality health care.
