An International Code of Conduct for artificial intelligence 2023/11/07

By: Juan Carlos Machorro

The G7 is an intergovernmental political forum made up of Germany, Canada, the United States, France, Italy, Japan and the United Kingdom that can be defined as an alliance made up of a select group of States whose structural positioning is similar based on the coincidence in their capabilities. national, without ideological barriers, with a willingness to coordinate policies towards the achievement of common objectives and the will to establish technical means of international cooperation.

On October 30, the G7 issued the Declaration of the International Hiroshima Process, related to artificial intelligence, which recognizes the opportunities for innovation and the transformative potential of artificial intelligence systems, notably foundational and intelligence models. artificial generative technology, and the need to adequately manage the risks inherent in its use, protecting people, society and the G7 principles regarding the rule of law and democratic values.

As a result of the G7's work on the matter, the so-called Code of Conduct of the International Hiroshima Process for the Development of Advanced Artificial Intelligence Systems has been issued, with the idea of ​​maintaining said document as a dynamic instrument that keeps pace with the technology and its constant evolution.

The objective is for the different companies and organizations that develop advanced artificial intelligence systems to commit to the application of this new code and accelerate the process of developing a structural framework for comprehensive policy on the matter by the end of this year in cooperation with the Global Partnership for Artificial Intelligence (GPAI) and the Organization for Economic Cooperation and Development (OECD).

This Code of Conduct constitutes a non-limiting list of actions and principles that, having taken into account the OECD's own AI principles, aim to take advantage of the benefits, consider and mitigate the risks, and challenges inherent to these technologies. It is a call for companies developing AI to adopt governance structures and policies and self-assessment mechanisms that facilitate a responsible approach to the implementation of these behaviors when developing AI.

In short, taking advantage of opportunities for innovation in the matter is of course encouraged, but, at the same time, organizations are invited to respect the rule of law, human rights, due process, diversity, equity, non-discrimination, democracy and, above all, maintaining the person as a primary value throughout the process of design, development and implementation of advanced artificial intelligence systems (AIS).

Specifically, the code includes the following eleven principles:

1. Take appropriate measures throughout the process of developing advanced artificial intelligence systems, including before starting said development, during it and when going to market, in order to identify, evaluate and mitigate risks throughout the development cycle. life of artificial intelligence.

2. Identify and mitigate vulnerabilities and, where applicable, incidents and, after implementation, including inclusion in the market.

3. Publicly disclose the capabilities, limitations and domains of advanced artificial intelligence systems, their appropriate and inappropriate use, in order to ensure the necessary transparency and increased responsibility and accountability.

4. Collaborate on the responsible exchange of information and incident reporting among those organizations that develop advanced artificial intelligence systems, including industry, governments, civil society and academia.

5. Develop, implement and disseminate the governance structure and risk management policies for artificial intelligence based on a risk measurement, including organizational policies and mitigation measures.

6. Invest in and implement robust security controls, including physical security, cybersecurity, and threats across the artificial intelligence lifecycle, including safeguards for insider threats.

7. Develop and implement reliable content authentication mechanisms and other techniques that allow users to identify artificially generated content.

8. Prioritize research to mitigate societal and security risks and prioritize investment in effective mitigation measures.

9. Prioritize the development of advanced artificial intelligence systems that address the most important global challenges, notably, but not limited to, the climate crisis, global health and education.

10. Advance the development and adoption of international technical standards.

11. Implement appropriate data entry measures and sufficient protections of personal data and intellectual property.

In the opinion of the G7, these measures, including the international code of conduct, will foster an open environment in which security prevails when designing, developing, implementing and using reliable artificial intelligence systems that allow maximizing the benefits of the technology while mitigating the risks. inherent to its use in protecting the global common good, including in developing and emerging economies with a view to closing the digital divide and achieving digital inclusion.