|The representatives of the 42 countries have affixed their signatures in a document that stressed how AI should be implemented and developed safely and fairly / metamorworks via Shutterstock|
A total of 42 countries have agreed on a set of guidelines regarding the use of artificial intelligence at the recent conference held in Paris. The first ever set of intergovernmental rules aims to ensure that artificial intelligence technologies are designed to be trustworthy, fair, safe, and robust. This is according to Reuters.
Global standards on how AI should be used and created
The representatives of the 42 countries, including the United States and Canada, have affixed their signatures in a four-page document that stressed how artificial intelligence should be implemented and developed safely and fairly. The global standards also call on technology firms to disclose information as to how their systems will work for the public so the people will understand what the tech’s result would be and how they can challenge them in case of adverse outcomes. The guidelines were created through an initiative spearheaded by the Organization for Economic Co-operation and Development, which is an intergovernmental economic organization founded in 1961 for the purpose of stimulating economic progress and world trade.
Comment from OECD’s Secretary-General about AI
OECD’s Secretary-General Jose Angel Gurria said during a meeting with the group members that although artificial intelligence has been driving optimism, it has also fuelled ethical concerns and anxieties among other people. For instance, there have been questions about the technology’s robustness and trustworthiness, such as the possibility of reinforcing or codifying biases related to race or gender or it could infringe on people’s right to privacy and basic human rights. The Mexican economist and diplomat further said during the official adoption of the AI guidelines that AI “is still in its infancy.”
Five principles for AI development as detailed in the agreement
There are five principles that are highlighted in the drafted guidelines, as well as five public policy recommendations. These principles include:
1. Inclusive growth and sustainable development and people’s well-being
This highlights that the stakeholders should engage proactively in “responsible stewardship” of AI in the hope that it will provide beneficial outcomes not only for the people but the planet as a whole. For example, it should be meant to enhance human creativity, augment their capabilities, and reduce gender, social, economic, and other forms of inequalities.
2. Human-centered fairness and values
This principle underlines that AI actors need to respect the law, democratic values, and human rights all throughout its system lifecycle. Data protection and privacy, autonomy and dignity, equality and non-discrimination, freedom, social justice, diversity, and fairness are some of the included rights and values.
3. Explainability and transparency
Artificial intelligence actors have to commit to responsible disclosure and transparency regarding their AI systems. This will help foster people’s general understanding of the technology and will allow those adversely affected by the technology to challenge the result based on easy and plain information from the AI actors themselves.
4. Safety, security, and robustness
The principle that gives emphasis on how AI systems need to be safe, secure, and robust all throughout its lifecycle in the hope that they will function appropriately.
The AI actors shall be accountable for the functioning of their technology and with respect to the law.
The AI guidelines may not be legally binding on the 42 signatories but are meant to influence the legislation of the governments. This is in relation to the fact that they have been facing pressure to provide a law as to what AI can really do and what are its limitations.
US government supporting the principles
The intergovernmental economic organization’s Director of the Science, Technology and Innovation Directorate Andrew Wyckoff likewise said via federal technology news provider NextGov that a part of the reason why they give so much value and importance to the drafting of the AI guidelines is that they recognize how AI is going to change almost every industry. Moreover, they have seen how it continuously affected the process of science and scientific discovery as well.
|Stakeholders should engage proactively in “responsible stewardship” of AI in the hope that it will provide beneficial outcomes not only for the people and the planet / SnvvSnvvSnvv via Shutterstock|
Reuters went on that although the administration of US President Donald Trump tends to pull out from various international agreements, including the North Atlantic Treaty Organization, the Washington officials supported the drafting of the AI guidelines. NATO is an international alliance consisting of 29 member states from Europe and North America. Part of the treaty states that in case of an armed attack against one of the member states, it shall also be considered as an attack against other members so the members will assist the attacked stated with armed forces.
The recently released international instrument titled Recommendation of the Council on Artificial Intelligence also indicates that their recommendation is open to non-OECD members and is calling for international cooperation. Included in the 42 countries that agreed on the AI guidelines are non-OECD members Romania, Peru, Costa Rica, Colombia, Brazil, and Argentina.
Prior to the drafting of the AI guidelines, the OECD has also set guidelines that became an international standard for issues like consumer protection and privacy law.