Algorithmic Accountability: Towards Accountable Systems

There is an ongoing move towards privatization of law enforcement online through algorithmic tools. This chapter discusses algorithmic accountability and its relevance for intermediary liability and human rights. First, the chapter looks into open issues related to the specific nature of accountability within the context of algorithmic accountability, especially regarding ‘to whom’ and ‘for what’ algorithms should be accountable. In doing so, the chapter considers algorithmic accountability to users, listing a number of technical, organizational, and regulatory challenges to make accountability possible in ensuring access to data. Considering intermediary liability and algorithmic accountability more closely, the chapter describes specific provisions for ensuring algorithmic accountability by online intermediaries and platforms, contextualizing them within a proposal in which adherence to algorithmic accountability would lower liability of intermediaries and contribute to more effectively ensuring compliance with human rights.

Maturity levels for private internal investigations

The purpose of private internal investigations by fraud examiners is to reconstruct the past by identifying past events and sequences of events. In this article, work by fraud examiners is studied in terms of maturity; introduce a five-stage model for investigation maturity: chaos, mess, disclosure, clarification, and investment. Based on student term papers in a financial crime class, six investigation reports are allocated to levels in the maturity model. The average score for the investigation reports is a level 3 disclosure, where the investigation is problem-oriented and often limited by the mandate. Based on the low average score, this article discusses the privatization of law enforcement, secrecy of investigation reports, lack of disclosure to the police, competence of private investigators, and limits of the investigation mandate.

A Nuanced Approach to the Privatization Debate

Current framing of the debate over the privatization of the State’s legislative and adjudicative functions masks the fact that there are distinct and conflicting versions of privatization of law. The different privatization models diverge on fundamental questions relating to the ontology of law, the role of social cooperation mechanisms in the lives of people, as well as the types of private legislative and adjudicative institutions that ought to replace the State’s legal system. In light of such conflicting normative premises, the distinct models of the privatization of law pose different kinds of challenges to both proponents and opponents of the privatization of law. At the outset, the Article juxtaposes two distinct visions regarding the privatization of law and adjudication—the market-based privatization model versus the community-based model. This analytical framework is then used to offer a fresh look at the privatization of law debate. The Article shows that the distinctions between the privatization models, especially with respect to the depiction of the social agents that are to replace the state’s lawmaking capacity, have great bearing on the advantages and disadvantages associated with privatization of law, and generate different types of costs and benefits. More specifically, the Article shows that the market-based model has greater susceptibility to market failures and to the under-provision of the public goods associated with the enterprise of law than the community-based paradigm. In addition, the market-based model runs a higher risk of corrupting the prevailing understanding of law as a collective, meaning-creating enterprise. The community-based model, on the other hand, has a greater vulnerability for coercion and is also more prone to political failures and public choice problems.