Accountability and transparency are two critical principles in AI value alignment as defined by IEEE Ethically Aligned Design guidelines. In some cases, for instance intelligent robots, their autonomous behaviors are not only decided by algorithmic autonomy but also the characteristics of physical embodiment. In this presentation, I would like to investigate the relevance between AI and embodiment in ethical value alignment.
Big data brings promises and perils. The US and EU have formulated various legal responses attempting to weed out the harm of big data analytics but harvest its benefits. The current legal concerns are largely on the harm of profiling, discrimination and misuse of personal data. This framework hinges on the identification of clear cases of data abuse and individual’s determination to seek legal redress. Scholars have advocated for data justice but without defining its content. In asking what a system of data justice requires to combat data abuse, China’s social credit system serves as a telling case study. This paper argues that the current legal frameworks are inadequate to cater the rising and all-encompassing dataveillance with ever evolving data analytics. A regime of data justice needs to address the increasing ties between public and private sectors, which data should not be re-combined and distributed, and how to achieve the aims of accessibility and accountability.
This paper discusses the legal implications of social media for the free exercise of the right to vote. In the last years, social media platforms (Twitter, Facebook) have been responsible for the rise and fall of several politicians. Social media platforms influence voters in numerous ways: they promote political debate, increase the electoral participation of young voters, organize voting polls, and select news articles for their users. However, the information of the public through social media is far from objective and transparent. Facebook employs obscure algorithms to shape the media exposure of users or disseminate information along certain voting lines. This paper discusses first the role of social media in enhancing democratic participation during electoral and non-electoral periods. Second, it analyzes the risk of ‘fake news’ during elections and the impact of social media on the right of voters to cast their ballot secretly, freely, and based upon an informed decision.
The National-level policy for the development of a Social Credit System (SCS) in the PRC has been laid out in a Planning Outline adopted by the State Council in 2014. Due to be fully implemented by 2020, the SCS consists of a socially-embedded informational platform for reciprocal reporting and decision-making based on the ascription of unified credit scores to people and institutions, encompassing all aspects of social life. By engendering a reflexive mechanism for social evaluation rooted in merit and grounding the conduct of public affairs in such a mechanism, the SCS has the potential to profoundly redefine China’s public sphere. This paper questions whether and how the SCS’s soft-law-based fabric might travel into Hong Kong’s legal system. Whereas the SCS is set to encompass “all information subjects … and all regions nationwide” – which may include the Hong Kong SAR and its residents – would this be compatible with Hong Kong’s constitutional order and the values that underpin it?
In December 2017, mainland China came up with a draft law providing comprehensive protection for national heroes and martyrs. The draft law prohibits twisting, vilifying, and negating the deeds and spirit of national heroes and martyrs. Acts infringing their names, images, reputation or honor can also give rise to civil liability, and be subject to administrative punishments or criminal sanctions. This draft law represents a further step in the protection of national heroes and martyrs because the PRC enacted legislation in March 2017 imposing civil liability on acts infringing their personality rights (except image right). Lobbyists for these recent initiatives have looked overseas for similar legislation and the Respect for Amercia’s Fallen Heroes Act has frequently been quoted. This presentation examines these recent legislative moves by the Chinese authorities and their impact on freedom of expression, with a view of comparing relevant American practices and scenarios.
This paper analyses the accountability problems in the deployment of big data in social governance by reviewing a core mechanism of the Social Credit System. The System subjects individuals to “joint punishments” that would substantially affect their interests across various fields of social lives, if their credit records process the traits determined by state datasets of social credit. The paper examines three factors that disrupt the existing mechanisms holding the government accountable for its decisions: (1) the privatisation and (2) semi-automation of decision making, and (3) the covertly imposed correlation of legally irrelevant factors. In exploring the approaches to addressing the accountability defects, it stresses the role of transparency as well as civic engagement pertaining to the quality of credit data, to the algorithm of credit rating, and to the relevance of credit records concerning different categories of behaviour to impose punishments legitimately.