The emergence of AI in all economic areas has put sectors such as healthcare on alert, which request that the new EU AI Law guarantee that AI systems marketed in Europe are safe and respectful of the fundamental values and rights of EU citizens and entities. To do this, they demand a specification related to health.
It must be taken into account that the new EU AI Law refers to essential services, but not to the health sector. For this reason, companies like ASHO have focused on the great benefits that AI will bring in terms of technical resources, precision, efficiency, and quality in patient care. However, it is also classified as ‘high risk’.
Ruth Cusco, managing director of LET’S SAY, has assured that during the coming years, the Healthcare sector will benefit from the application of AI tools that will improve, from a technical and direct patient care perspective. Therefore, it is necessary to establish clear and rigorous regulations in this regard.
Julius Clement, director of ASHO’s Artificial Intelligence area, believes that AI will help streamline and reduce the time spent collecting and interpreting data, automating everyday tasks. In addition, it will ensure the offering of personalized treatments that can quickly change the way medicines are made.
Thus, if there is a European body that ensures compliance and updating of AI risks, it will be considered which actions and advances pose a risk to science and which represent support. All this, taking into account the creation of a consultative forum through society and SMEs who will be able to freely express their concerns and interests.
AI in the healthcare sector can be applied in the collection of diagnoses and clinical care, in the investigation of symptoms and treatments, in daily and repetitive administrative tasks, in medical and nursing education, and in scientific research and drug development and new compounds.
Nevertheless, WHO warns about the need to prevent risks of false, inaccurate, or biased statements. Additionally, LLMs may be trained on low-quality data or biased by race, sex, ethnicity, ancestry, age, or gender identity.
The EU intends for the new legislation on AI to be in force internationally. Although it will come into force, according to estimates, in 2026, it is necessary to begin work on the structures necessary for its compliance in Europe and the rest of the world. Those companies that violate the regulations and depending on their size and the type of inflation, they could face fines that will range between 35 million euros or 7% of global turnover, up to 7.5 million euros or 1.5% of overall billing.
Recently, the World Health Organization (WHO) has publicly expressed, through a statement, its position regarding the application of AI in the healthcare sector and, specifically, on the limitations that the new EU AI Law should impose.
According to the WHO, it will be essential to include healthcare professionals and patients in the early stages of developing LLM applications and Generative AI technology for the healthcare sector. You should therefore bet on structured and fully inclusive and transparent designs.
It is advocated because there exists collaborative work between healthcare professionals and developers, capable of addressing ethical issues and providing essential information for creating the best AI in the healthcare sector. Medical providers and scientific researchers should also be present in this process.
Specifically, the WHO has developed a guide made up of more than 40 recommendations for Governments on AI. Based on this, Jeremy Farrar, chief scientist at the WHO, points out that Generative AI has sufficient potential to improve healthcare as long as its developers identify the associated risks. It is necessary to establish 100% transparent information and policies to manage the design, development, and use of LLM to achieve better outcomes and overcome all persistent health inequalities.