Sunday, June 07, 2020


Somehow, this does not give me confidence that Facebook is ready for our next election.
Operation Carthage: How a Tunisian company conducted influence operations in African presidential elections
A Tunisia-based company operated a sophisticated digital campaign involving multiple social media platforms and websites in an attempt to influence the country’s 2019 presidential election, as well as other recent elections in Africa. In an exclusive investigation that began in September 2019, the DFRLab uncovered dozens of online assets with connections to Tunisian digital communications firm UReputation. On June 5, 2020, after conducting its own investigation, Facebook announced it had taken down more than 900 assets affiliated with the UReputation operation, including 182 user accounts, 446 pages, and 96 groups, as well as 209 Instagram accounts. The operation also involved the publication of multiple Francophone websites, some going back more than five years.
In a statement, a Facebook spokesperson said that these assets were removed for violating the company’s policy against foreign interference, which is coordinated inauthentic behavior on behalf of a foreign entity. “The individuals behind this activity used fake accounts to masquerade as locals in countries they targeted, post and like their own content, drive people to off-platform sites, and manage groups and pages posing as independent news entities,” the spokesperson said. “Some of these pages engaged in abusive audience building tactics changing their focus from non-political to political themes, including substantial name and admin changes over time.”




Rather gloomy…
On the Ethics of Artificial Intelligence
Currently, AI ethics is failing in many cases. Ethics lacks a reinforcement mechanism. Deviations from the various codes of ethics have no consequences. And in cases where ethics is integrated into institutions, it mainly serves as a marketing strategy. Furthermore, empirical experiments show that reading ethics guidelines has no significant influence on the decision- making of software developers. It is a boom time for artificial intelligence (AI) and ethics. All sorts of groups have launched manifestos, declarations, toolkits and lists of principles to set the ethical agenda. There are so many lists of principles that now other groups are providing guides to all the lists. You would think that having all these principles and checklists would be a good thing, but many of them are being generated by industry or by scientists. We risk ignoring other approaches to ethics that come from the humanities. In this panel we will present a dialogue of philosophical perspectives and informatics approaches on artificial intelligence (AI). These reflect an interdisciplinary collaboration at the University of Alberta between faculty and students across Digital Humanities, Philosophy, Communications, and Library and Information Studies




The tools already exist?
A “right to explanation” for algorithmic decisions?
In discussions about the legal hurdles facing the increased use of artificial intelligence (AI), academic debate has focused on civil and criminal liability for damage caused by AI. At the same time, however, data protection law also creates challenges to the increased use of intelligent systems and machines, i.e. machines which are capable of learning, and these challenges should not be underestimated. As a means of protecting fundamental rights1, data protection law is used to safeguard each individual’s general right of personality, and in particular, his right to determine what information concerning him is made available or known to parties in his surrounding environment.2 To this end, the law has in its arsenal procedures, mechanisms and rights, which apply to every processing of personal data within its scope, even if the data is handled not by a human processor but by a self-learning system.




Perspective. In China, Privacy is not about the ‘personal?’
Personal information Legislation in the Age of Big Data and Artificial Intelligence: Challenges of New Technology to Information Privacy
Abstract: The development of new technology has posed new challenges to the traditional framework of data protection. In the context of big data, artificial intelligence, internet of things, cloud technology and so on, authorization of personal information based on user consent often falls into invalid or hinders technological progress, and meanwhile, cannot really provide data privacy protection. Besides, there are also problems with some principles of restricting data users. The reason is that the current personal information protection law adopts the individualism and static protection approach, and there is tension with data utilization under the background of new science and technology. In the future, the protection of personal information should shift from individualistic permission to risk-oriented public law control, from static protection to dynamic protection.




Perspective. Amazon is probably not alone.
Amazon’s Heavily Automated HR Leaves Workers in Sick-Leave Limbo
The company is struggling to handle thousands of requests from ailing employees and those who must stay home to care for kids or elderly relatives.
the design of Amazon’s HR department reflects the strengths and weaknesses of the company’s culture. It’s heavily automated, which helps Amazon grow quickly and restrain costs but these days leaves employees hitting dead ends with chatbots, smartphone apps and phone trees.
Three people with experience in the company’s human resources group say the unit has been weighed down by competing priorities. HR is expected to offer workers the same speedy customer service as Amazon’s customers, while practicing a level of frugality that Amazon sometimes takes to extremes, the employees say.
HR “is always struggling to automate and keep pace with the scale of the company,” says one of the people, who all requested anonymity because they signed confidentiality agreements. “The horror stories happen because [HR] people are overwhelmed. And they don’t have the resources and the mental capacity to deal with [workers] because they’re pulled in so many different directions. It’s bound to have negative, real-life human impacts.”



No comments: