Formulating policy options for big data and AI development

27 Nov 2019 09:30h - 10:30h

Event report

[Read more session reports and updates from the 14th Internet Governance Forum]

Artificial intelligence (AI) has provided a significant number of benefits to our lives but has also caused disruptions in our economy, labour market, and human rights. The conference discussed the way forward to developing national and international policy on AI by applying UNESCO’s ROAM principles of: Human Rights, Openness, Accessibility, and Multistakeholder governance) developed to use AI and other technologies to achieve the sustainable development goals (SDGs).

UNESCO published the report ‘Steering AI and Advanced ICTs for Knowledge Societies,’ which explains the positive developments brought by AI, as well as its potential risks and challenges. The report analyses the impact of AI influencing various aspects of our lives. On the one hand, the right to information has been improved due to better Internet connectivity and proliferation of platforms. On the other, the exact same right could potentially be hindered by AI because it selects news and information that we see on the Internet based on our data. In regard to freedom of speech, AI is now removing hate speech and harmful content from online space, however, we need to ask ourselves what that means to freedom of speech. It is even considered problematic because of algorithmic biases, Ms Xianhong Hu (UNESCO) noted that we need to work towards eliminating bias in algorithms as the prevalence of automated decision-making has exponentially increased. Ms Jai Vipra (Tech Policy Researcher, IT for Change) explained that it is possible to reduce algorithm biases by regulating AI outputs and AI-generated decisions, whilst it is quite challenging to control the inputs, algorithms.

Disinformation and its impacts on electoral processes was another highlight from the session. While technological developments in general contributed to the improved system in elections by enabling e-voting, we have seen the use of AI for disinformation, deep fakes, and voter suppression. With these harmful activities, the use of information and communications technology (ICT) can potentially undermine the trust and credibility towards the democratic electoral systems. Mr Robert Krimmer (Managing Director and Founder, E-Voting.CC) pointed out the importance of regulation and policy in this context because freedom of speech could sometimes be the dilemma of democracy with or without technology.

The panel also highlighted the two tensions between access to information, education, science, and culture. While AI can contribute to building an inclusive knowledge society, it can also accelerate the inequality of access to quality education, science, and culture between the wealthy population and the poor. Mr Alex Comninos (ICT Policy Researcher, Research ICT Africa) stressed the importance of reducing the digital divide across the globe simultaneously as Internet connectivity is a prerequisite for all the benefits generated by AI.

The reactionary tendency of policymakers towards AI issues was pointed out by the participants. To be proactive in a regulatory setting, it is integral to promote the multistakeholder approach, since the expertise on AI and other technologies mainly resides in the private sector. Moreover, to make AI a human-centric tool, the responsibility of the private sector was also noted as it can ensure the potential impact and implications of their products on human rights, before they release them to the market, or even before they develop the products.

By Nagisa Miyachi