The AI Asia Pacific Institute contributed to the UNESCO online public consultation on the ethics of artificial intelligence. The initiative is part of a mission to provide everyone with the opportunity to participate in the work of its international group of experts on AI.
We compliment UNESCO for the initiative in establishing the Project. Particularly for recognising that addressing risks and ethical concerns is the right path towards stimulating new practices of responsible research and innovation. We wish to highlight the following values and principles which make part of the first version of the Recommendation’s draft text (the Recommendation).
The Recommendation encourages a certification mechanism for AI systems similar to the ones used for medical devices. UNESCO suggests that this can include different classes of certification according to the sensitivity of the application domain and expected impact on human lives, the environment, ethical considerations such as equality, diversity and cultural values, among others. Most importantly, the Recommendation suggests that such a mechanism must not hinder innovation or disadvantage small enterprises or startups by requiring large amounts of paperwork.
Recommendation 58 encourages the adoption of the role of an AI Ethics Officer who would be responsible for overseeing impact assessment, auditing and monitoring efforts with the aim of ensuring ethical compliance of AI systems. This suggestion has received a lot of support from different experts recently. Kay Firth-Butterfield, Head of Artificial Intelligence and Machine Learning, World Economic Forum in her recent article to Forbes describes how such role could become crucial in addressing questions around the safety, privacy, accountability transparency and bias in algorithms.
The Recommendation also encourages the development of data governance strategies that ensure the continuous evaluation of the quality of training data for AI systems.
Importantly, promoting AI ethics research has been considered as one of the areas of policy action under the Recommendation. AI researchers are to be trained in research ethics and encouraged to include ethical considerations in their research design and end products.
Most critically, UNESCO recommends for Member States to consider the development and implementation of an international legal framework to encourage international cooperation between States and other stakeholders.
Emerging technologies, including artificial intelligence, are transnational and so the challenges arising from them should be considered to affect the international community as a whole. In this context, investing in avenues for international dialogue on AI is crucial in ensuring that its development is in line with what we want.
We are confident the UNESCO Recommendation will play a role in defining shared values and principles, and in identifying concrete policy measures on the ethics of AI. Our Institute remains committed to supporting this goal and in working collaboratively on a strategy to develop trustworthy AI.