As part of growing NLP capabilities, coupled with an awareness of the ethical dimensions of research, questions have been raised about whether particular datasets and tasks should be deemed off-limits for NLP research. We examine this question with respect to a paper on automatic legal sentencing from EMNLP 2019 which was a source of some debate, in asking whether the paper should have been allowed to be published, who should have been charged with making such a decision, and on what basis. We focus in particular on the role of data statements in ethically assessing research, but also discuss the topic of dual use, and examine the outcomes of similar debates in other scientific disciplines.

A key part of the NLP ethics movement is responsible use of data, but exactly what that means or how it can be best achieved remain unclear. This position paper discusses the core legal and ethical principles for collection and sharing of textual data, and the tensions between them.We propose a potential checklist for responsible data (re-)use that could both standardise the peer review of conference submissions, as well as enable a more in-depth view of published research across the community. Our proposal aims to contribute to the development of a consistent standard for data (re-)use, embraced across NLP conferences.

Privacy protection legislation and policy is heavily dependent on the notion of de-identification. Repeated examples of its failure in real-world use have had little impact on the popularity of its usage in policy and legislation. In this paper we will examine some of the misconceptions that have occurred to attempt to explain why, in spite of all the evidence, we continue to rely on a technique that has been shown not to work, and further, which is purported to protect privacy when it clearly does not. With a particular focus on Australia, we shall look at how misconceptions regarding de-identification are perpetuated. We highlight that continuing to discuss the fiction of de-identified data as a form of privacy actively undermines privacy and privacy norms. Further, we note that ‘de-identification of data’ should not be presented as a form of privacy protection by policy makers, and that greater legislative protections of privacy are urgently needed given the volumes of data being collected, connected and mined.

No abstract

T his year, on 17 June, the Geneva Protocol, an international treaty prohibiting the use of asphyxiating or poisonous gases and bacteriological methods of warfare, turned 80 years of age. It was fostered in part by a 1918 appeal in which the International Committee of the Red Cross (ICRC) described the use of poisonous gas against soldiers as a "barbarous invention which science is bringing to perfection." Great peacetime advances in chemistry before the First World War made possible the manufacture and use of chlorine, phosgene, and mustard gas on the battlefield. The ICRC foresaw "a struggle the ferocity of which will exceed the greatest barbarity the world has known." The 20th century witnessed a continuous accumulation of potential biological and chemical weapons in many nations. Some of these weapons were deployed; for example, in Abyssinia in the 1930s, in China in World War II, in Yemen in 1963, and in the Iran-Iraq war of the 1980s. This gave the impetus to two important international legal regimes: the 1972 Biological Weapons Convention (BWC) and the 1993 Chemical Weapons Convention (CWC). These treaties extended the Geneva Protocol's prohibition beyond mere use to include the development, production, and transfer of all such weapons. The scientific community now faces difficult questions. Major advances in chemistry, microbiology, and nuclear physics have, regrettably, led to hostile use of the knowledge and materials from these scientific domains; a use that the original scientistdiscoverers would have deplored. What will be the outcome for humanity if the results of the research explosion in life sciences and biotechnology are also turned to hostile use? What are the associated responsibilities of scientists? Scientists in academia and government recognize that advances in the life sciences and biotechnology could make biological weapons more effective, safer to use, more difficult to detect, and therefore more attractive options for would-be users. The dangers lie in both bioterrorism and in government-backed programs, and robust international mechanisms are required to deter the development, production, transfer, and use of these weapons. This provides the background to the numerous conferences this year discussing biosecurity and for the meeting of experts at the BWC this month in Geneva that is examining proposed codes of conduct for scientists. These discussions take place against a history of societal norms against biological and chemical weapons that reaches back much further than the 1925 Geneva Protocol. The current treaties represent the development and codification of rules and taboos that for thousands of years have protected people from poisoning and the deliberate spreading of disease. Both Greek and Roman civilizations customarily observed a prohibition on the use of poisons. In 500 B.C., the Manu treaty in India banned such weapons. A millennium later, regulations on the conduct of war drawn from the Koran by the Saracens forbade poisoning. These examples demonstrate that recent treati...