[Read more session reports and live updates from the EuroDig 2019]
The session discussed whether artificial intelligence (AI) is related to Internet governance. It questioned whether AI is a service or a service buried in a product. It also questioned whether AI uses the Internet or the Internet itself is AI.
It was stated that there is a lack of clarity about what is meant under the term AI. A point was made that there is a whole field of AI outside of the Internet. Another view expressed was that AI obtains its information through the Internet – the Internet is the source of data that is fed into AI. Also discussed was the topic of how AI will use the Internet to communicate with other AI. One of the views expressed was that the network itself can become intelligent – an AI.
Another opinion stated was that AI will have a heavy effect on economic and ethical issues, and that the dimension of the problems it could cause is so complex humans cannot afford to let it go unregulated. In this regard, Internet governance is important because it operates in a multistakeholder model that showcases how to deal with technologies where problems are at a global scale, complex, and various, such as the issues AI will raise.
AI as a tool for Internet governance was debated. If AI is the solution to Internet governance problems, then a set of rules for AI has to be decided. The Internet Governance Forum (IGF) was suggested as a venue for AI governance discussions. On the other hand, the International Telecommunications Union (ITU) was also suggested as a venue for discussing emerging technologies. It was also stated that the governance of AI is too broad because what is considered AI is too broad itself.
AI was also discussed in relation to content moderation, more specifically with regard to hate speech. Here it was stated that the problems of the Internet are at such a scale that they are not solvable by humans, and that AI is needed to solve them.
The question of access to AI was brought up. There is a concern that countries with poor access could be left behind when it comes to AI development and use.
By Andrijana Gavrilović