AI and ethics in health: a complicated binomial to manage

The 11th R + D + I Tic Social Health Conference was inaugurated with the participation of Ignacio Fuentes, a researcher at the Massachusetts Institute of Technology (MIT), Eleonora Harwich, head of collaborations in the Public Health System at Intel’s Lab UK Artificial Intelligence; and Joan Carles March, professor at the Andalusian School of Public Health.

(Photo: Ignacio Fuentes and Karma Peiró, at the ICT Social Health Conference 2021)

Fuentes stressed the importance of regulation in Artificial Intelligence solutions, as well as the difficult “access to data that is one of the main barriers to achieving robust models.” He also presented the main objectives of the Jameel Clinic – MIT’s Center for Research in AI and Medicine – and some of the projects under development.

Harwich then explained how AI has entered the British health system, detailing the different programs and projects in which they work, with examples of possible applications already today in the field of health. The researcher stressed that ethical aspects and current European regulation are always present.

March shared reflections from the point of view of patients, of the end users of these services and projects that are being developed. In his speech, March recalled the human values ​​”that must be preserved in this different and technological 21st century.”

(Photo: from left to right, Eleonora Harwich, Karma Peiró and Joan Carles March)

The debate between Harwich and March highlighted the importance of achieving a common and cooperative Artificial Intelligence strategy. “It’s not an easy challenge,” the NHS AI Lab representative clarified. “We are precisely defining guidelines and guidelines for understanding the direction in which we need to move forward, setting priorities and respecting regulation in this regard,” he added.

Another aspect discussed was the role of citizens in defining the ethical foundations of the evolution of AI, both in its application to the health system and in general. In this sense, March considers essential “the validation and participation of citizens.” And it proposes to train patients on the basic elements of bioethics, so that they can intervene in conversations with professionals and contribute more nuances to the design and development of this technology.

The session concluded with reflections and questions around concepts such as transparency, integrity and justice.

(See the full 1st session at this link)

What happens when ethics comes into consideration?

The session on October 21 was dedicated to the binomial AI + ethics. And the complexity increased because it was no longer a matter of presenting only projects but of talking about concepts that must serve to respect acquired human rights combined with technological innovation. The guest speakers on this occasion were Eduard Fosch, assistant professor at Leiden University; Albert Sabater, director of the Ethics Observatory of the Catalan Government; and Karina Gibert, Intelligent Data Science and Artificial Intelligence researcher.

(Photo: from left to right, Karma Peiró, Karina Gibert, Eduard Fosch, Albert Sabater)

Fosch focused on the importance of integrating gender and sexuality into the development of AI tools applied to health. The researcher explained the implications that can be had if the detection of biases is not taken into account and diversity considerations are ignored “while building what is called the normality of the population”. How AI-based systems can reinforce biases in healthcare, with legal consequences if not addressed properly. Fosch’s research seeks to understand the risks and opportunities that robots and AI pose in the legal field.

Gibert presented four projects in which artificial intelligence has helped to take better solutions in complex phenomena. In all cases, the ethical challenges posed by AI are revealed. The critical points are always the processing of personal data, the identification of people, the ability to personalize responses with anonymous data and the impossibility of implementing human supervision without distorting the system. According to Gibert, for all these challenges “the European vision is the strictest on a global scale, focused on the person and oriented to the common good.”

In his speech, Sabater emphasized the need to apply ART – an acronym for ethical concepts such as Accountability, Accountability and Transparency – to artificial intelligence solutions. One of the highlights was the distinction between “what is morally acceptable and what is legally allowed, so that in this transformation we do not lose the characteristics that define us as human beings, such as empathy and ethics.” remarked Sabater.

During the discussion between the speakers, the roles and responsibilities were discussed. From individual responsibility to corporate responsibility and that of public legislators. In this context, citizen responsibility is growing and it is necessary to accompany the citizenry to prevent existing social inequalities from being transferred to the use of AI. He also delved into the role of Europe, where it seems that the path he is taking is the quality of AI from an ethical point of view.

In conclusion, it was important to think long-term and regulate the principles on which it should be based so that it becomes a fair tool for all.

The XI Conference R + D + I Tic Social Health 2021 –celebrated online during the days 20, 21, 27 and 28 October–, were organized by the Chair TIC Social Health of the University of Vic – Central University of Catalonia (UVic-UCC), the TIC Social Health Foundation (Department of Health of the Generalitat de Catalunya), the Vic Hospital Consortium, the Santa Creu de Vic Hospital Foundation, the EAP / ABS Centelles, the Vic EAP, the Catalan Institute of Health-Management Central Catalonia, Althaia Foundation and FORES. (See full 2nd session at this link)