Participation in the consultation of: United Nations Advisory Committee Interim Report on AI Governance for Humanity.
Provide a brief description of your organization/entity
The CSAIA Association involves several Italian universities, the National Research Centre, and many important Italian and international industries exploring all fields of human knowledge, combining the humanistic aspect of the issue with law, philosophy, the arts, labour law, together with medicine, engineering, physics, mathematics, cybersecurity, aerospace, and agriculture. We believe that AI studies and applications can emerge from a collaborative path between the public and private sectors.
Opportunities and enabling factors
Considering that AI represents a true new industrial revolution in all fields of everyday human knowledge, AI could change the classical way of accessing knowledge by providing the real light for a new revolutionary approach.
AI is capable of increasing productivity in all aspects of traditional human knowledge, from science, medicine, engineering, energy, and environmental sustainability to the legislative process for creating rules for AI governance, as well as aerospace, automotive, and agriculture.
Artificial intelligence has many tools to support the traditional educational method. Applications have been developed to act as virtual assistants for many people not only through translation programs, but also through new generative AI.
Any aspect of human knowledge can become revolutionary in gaining new perspectives. In medicine, AI-based tools that predict protein structures are used by many researchers for drug discovery and to advance the understanding of diseases, including many previously neglected diseases. Artificial intelligence is powering diagnostic tools to help physicians detect various types of diseases more promptly, thereby saving human lives. In the energy sector, AI is playing a role in optimizing energy systems and advancing the transition toward renewable energy sources. In order to prevent AI from being regarded as a tool against the Human Being, it is essential to strive for the achievement of global governance capable of promoting all potential objectives.
Risks and challenges
All new revolutionary technologies may carry the risk of causing significant harm to human society. In relation to AI systems, risks may arise both from technical limitations of the systems (i.e. biases and hallucinations of generative AI) and from their application and use.
Naturally, it is essential to have a group of experts capable of detecting every risk for the human being as an entity and as part of the community.
AI must avoid deception, nudging, biometric recognition, as well as excessive use of and dependence on technology, gender discrimination, and any violation of ethical values and strict legal rules.
From these considerations, an international AI governance framework should be created through which guidelines for the application of AI could be identified and enforced worldwide, independently from economic power pushing AI Systems not guided by human beings.
Human beings have the role of creating the path, the rules, and the appropriate applications of artificial intelligence systems.
Guiding principles for shaping the creation of new global AI governance institutions
Guidelines for the application and use of AI should include several points:
- Transparency: it should be ensured to allow all stakeholders to fully understand the application of AI systems, encouraging companies to optimize monitoring practices.
- Inclusion: AI systems must not discriminate against any human being because of equal dignity, and all citizens should have the opportunity to improve the quality of their lives.
- Responsibility: every investigation, from the design phase to the research application of the artificial intelligence system, must respond to the required task, and every AI company must involve not only the company itself but also universities, small private AI laboratories, and other organizations dedicated to artificial intelligence.
- Impartiality: artificial intelligence systems, thanks to human oversight, must not create any bias.
- Reliability: the artificial intelligence system must be reliable for every human being.
- Security and privacy: following the directives required by the Authorities, governance must be created to ensure the proper protection of privacy.
All the above-mentioned points may be integrated into a common framework of AI principles, guidelines, and governance, including the European Parliament’s AI Act (currently under evaluation for regulation by the European Commission), the United Nations Charter on AI Governance, and international human rights law.
Institutional functions that an international AI governance regime should perform
AI governance requires several interconnected lines of action.
International organizations, governments, private companies, and academia should have different roles but pursue the same objectives for these functions.
The application of AI requires consensus among all involved actors, and citizens are not the final step, but rather the most important actors to involve for the correct understanding and use of AI Systems.
Governance could become the independent branch responsible for monitoring the trajectory and evolution of AI systems, developing and harmonizing safety and risk management, improving economic conditions, promoting international cooperation, and pooling expert knowledge.
From this perspective, strict legal rules are mandatory not only to assess changes, but also to emphasize both transparency and accountability among all actors involved.
Additional comments on the International AI Governance section (apart from Principles and Functions, addressed in previous questions)
International governance should be:
Transparent
Inclusive
Accountable
Impartial
Reliable
Respectful of the human being at the center of every AI System process
Any further feedback on the interim report
As President of the CSAIA Association, I believe that the Interim Report fully reflects all the circumstances under which it was created. AI systems will change every aspect of human life within a few years and, without rules, cooperation, and consensus, AI could become dangerous for humanity itself.
The CSAIA Association would be honored to take part, together with its researchers and institutional stakeholders, in this immense challenge for humanity. We are ready to cooperate in creating consensus on the rules to be applied.
