About the Author: Peter G. Kirchschlaeger is a Full Professor of Theological Ethics and Director of the Institute of Social Ethics ISE at the University of Lucerne, Research Fellow at the University of the Free State, Bloemfontein (South Africa), Visiting Professor at the Chair for Neuroinformatics and Neural Systems at ETH Zurich as well at the ETH AI Center (one of the world's largest hubs for research in the field of AI), and Visiting Fellow at the University of Tuebingen (Germany) which is closely linked with the Cyber Valley– Europe’s largest and leading center for excellence in AI and modern robotics. Prior, he was Visiting Fellow at Yale University (USA).
In his research, he focuses on the ethics of AI and digital transformation, human rights ethics, and business ethics. He serves as a consultative expert on ethics for international organizations, institutions, and companies, and is, among other roles, the President of the Swiss Federal Ethics Committee on Non-Human Biotechnology. His latest books include: “Ethical Decision-Making” (2023) and “Digital Transformation and Ethics. Ethical Considerations on the Robotization and Automation of Society and the Economy and the Use of Artificial Intelligence” (2021).
Artificial Intelligence and the Rights of the Child
From an ethical perspective on children of today and children of tomorrow, traditional artificial intelligence (“AI”) and generative AI have profound positive and negative effects. AI affects, among other facets, human thinking and interaction, sciences and culture, education, information, and communication. Using the rights of the child as an ethical frame of reference – which is protecting, for example, physical and mental integrity, freedom, autonomy, non-discrimination, and security – allows us to precisely identify the ethical upsides and downsides of AI Too often, academic discourse on AI and the rights of children creates the following misconception: As the rights of the child are at stake, every single violation of children’s rights is ethically unacceptable. A violation of children's rights with or through AI cannot be outweighed by the positive ethical potential of AI in terms of children's rights. However, the following examples illustrate the positives and negatives of AI on children’s rights.
For example, AI fosters a child's right – the right to education especially in this case – when AI and generative AI in the form of an individualized learning assistant are supporting children’s learning, development, and communicative skills. At the same time, violations of children’s rights to privacy and data protection are occurring because of these activities specifically in “smart toys” which generate and collect the data of children.
The rights of the child are further under attack because AI and generative AI lead to an enormous increase of already existing and emerging sexually exploitative practices against children in the digital environment. This causes a massive negative impact on children’s right to physical and mental integrity.
Moreover, the negative impact of AI and generative AI on the mental health of children is a violation of the children’s right to physical and mental integrity.
With specific channels for children and with the rhythmizing of its programs, YouTube Kids and others strive for creating the obsession of watching as many videos as possible. Beyond that, targeting children based on machine learning algorithms (e.g., on YouTube and YouTube Kids) and collecting their personal data is violating children’s rights to privacy and data-protection. This enormous amount of data – which allows to know children better as they know themselves – can then be used for political manipulation (forcing children to think what one wants them to think politically) and economic manipulation of children (by pushing children to consume what one wants them to buy) online representing violations of the freedom and autonomy of children. Furthermore, age-inappropriate content is just a few clicks away or part of from automatic advertisements interrupting the transmissions. This lack of online security protecting children violates the children’s right to security.
Furthermore, algorithms and data are neither objective, neutral, nor fair. Algorithms and data, including decisions and actions by AI, have biases, for example, related to gender, skin color. These biases cause discrimination and exclusion. For example, an AI-learning assistant would evaluate due to its biases the performance of a child differently depending on the gender or the skin color of the child. This violates the children’s right to non-discrimination.
In addition, the exploding digital divide and the severe ecological exploitation by this technological progress due to its carbon impact contributing to climate destruction. Also, the excavation of the necessary natural resources and the production of the technological applications destroy the environment and possess a carbon-footprint. Poverty as a consequence of the digital divide and climate destruction in turn affect children’s suffering.
the main goal of the use of AI consists in the strive for efficiency by, among other factors, reducing costs through the replacement of humans (e.g., with automated check out in grocery stores);
AI is made on purpose for the goal that always less to no human input will be needed;
AI impacts in this way all professional tasks – not just the ones who require low or any qualification.
This impact of AI on human labor causing a massive reduction of paid professional tasks affects the rights of the child indirectly as they experience the negative socio-economic consequences as well as directly because only a minority of them will ever fulfill a paid professional task.
As it stands today, the rights of the child are neither sufficiently respected, protected, implemented, nor realized in the field of AI. United Nations (UN) Secretary-General António Guterres pointed out in the UN Security Council on July 18, 2023: “Generative AI has enormous potential for good and evil at scale... Without action to address these risks, we are derelict in our responsibilities to present and future generations.” It represents the responsibility of humans to take urgent action ensuring that the ethical upsides of AI will be achieved and the ethical downsides will be avoided or mastered in order to allow the children of today and of tomorrow to flourish on a planet with a sustainable future.
AI? Data-Based Systems (DS)
A first step of living up to this human ethical responsibility for AI and to address the ethical dimension of AI embraces the critical examination of the term AI. From an ethical perspective, in view of the nature of AI, doubts arise as to whether the term is even adequate. AI strives to imitate human intelligence, but this is limited to a certain area of intelligence (e.g., certain cognitive capacities). Among other areas of intelligence, in the domain of emotional and social intelligence, machines can only simulate emotions, personal interactions and relationships, and therefore lack authenticity. For instance, one can train a healthcare robot to cry when a child patient cries, but no one would argue that the robot feels real emotions and cries due to them. On the contrary, one could train the same robot to slap the child patient’s face when the patient is crying, and the robot would perform this function perfectly. Beyond that, the healthcare robot cannot set ethical principles and norms providing normative guidance that slapping a child isn’t appropriate because of the lack of moral capability. One cannot ascribe machines with moral capability because they are presupposed by patterns and rules defined by humans and do not possess vulnerability, conscience, freedom, responsibility, and autonomy. Therefore, the ethical quality of its action is not accessible to a healthcare robot. Humans need to train machines in ethical principles and norms revealing then the illegitimacy of this violation of the physical and mental integrity of the child patient.
The term “data-based systems (DS)” would be more appropriate than “artificial intelligence” because the former describes what constitutes “artificial intelligence”: generation, collection, and evaluation of data; data-based perception; data-based predictions; and data-based decisions.
Recognizing and acknowledging the core characteristic of DS allow more accurateness, adequacy, and precision in the critical reflection on DS. At the same time, humans possess the exclusive responsibility to identify ethical opportunities and risks of DS precisely to allow children of today and tomorrow to flourish with their human dignity respected on a planet with a sustainable future. To guarantee the respect, protection, implementation, and realization of the rights of the child globally online and offline, in the domain of DS, the following concrete measures should be realized.
Rights of the Child-Based Data-Based Systems (RCBDS)
Rights of the child-based data-based systems (RCBDS) should ensure that the rights of the child serve as the basis of DS. In other words, RCBDS seeks to ensure that states and the private sector respect, protect, implement, and realize the rights of the child within the entire life cycle and value-chain process of DS (in the design, development, production, distribution, use, or non-use of DS because of rights of the child-concerns). RCBDS also embraces the necessity to increase children’s participation in the opinion-forming and decision-making processes to guarantee the rights of the child.
International Data-Based Systems Agency (IDA) at the UN
An International Data-Based Systems Agency (IDA) must urgently establish at the UN as a global platform for technical cooperation in the field of DS. The IDA should serve to foster human rights including rights of the child, sustainability, safety, security, and peaceful uses of DS. The IDA should act as a global supervisory and monitoring institution, and as a regulatory authority in the area of DS responsible for access to market approval.
Given the areas of convergence between DS and nuclear technologies both possessing ethical upsides and downsides, the International Atomic Energy Agency (IAEA) model would seem the most appropriate for responsible global AI governance as this model represents a UN agency with “teeth” because of its ability to enforce a regulatory framework respectively to sanction misconduct.
There is a growing international and interdisciplinary network of experts calling for the establishment of RCBDS and IDA. “The Elders,” an independent group of world leaders founded by Nelson Mandela, have endorsed RCBDS and a global agency like IDA to monitor DS. The ideas of a human rights-based and legally binding regulatory framework and an institution enforcing global regulation of DS enjoy the support of Pope Francis. UN Secretary-General Guterres also supports the creation of an international AI-watchdog-body like the International Atomic Energy Agency (IAEA): “I would be favorable to the idea that we could have an artificial intelligence agency … inspired by what the international agency of atomic energy is today.” He has called for a new UN body like an “International Data-Based Systems Agency IDA” to tackle threats posed by artificial intelligence in the UN Security Council on July 18, 2023.
On July 14, 2023, the UN Human Rights Council unanimously adopted its latest resolution on “New and emerging digital technologies and human rights” which included for the first time an explicit reference to AI as it relates to promoting and protecting human rights. The resolution emphasized that new and emerging technologies with an impact on human rights “may lack adequate regulation”. The resolution highlighted the “need for effective measures to prevent, mitigate and remedy adverse human rights impacts of such technologies”. It stressed the need to respect, protect, and promote human rights “throughout the lifecycle of artificial intelligence systems.” The resolution called for frameworks for impact assessments related to human rights; due diligence to assess, prevent, and mitigate adverse human rights impact; and effective remedies, human oversight, and accountability. In other words, from a human rights perspective including the rights of the child, DS impact human rights including the rights of the child. Therefore, the design, development, and production, and use of DS must respect of human rights and rights of the child.
On March 21, 2024, the UN General Assembly unanimously adopted a resolution, “Seizing the opportunities of safe, secure and trustworthy artificial intelligence systems for sustainable development.” The resolution emphasized: “The same rights that people have offline must also be protected online, including throughout the life cycle of artificial intelligence systems.” Now, it is urgent for all UN member states together with the private sector to implement and build on the resolution of the UN General Assembly to live up to their responsibility to present and future children.
Comments