Miriam Fahimi im Standard-Interview über Gefahren von Künstlicher Intelligenz

/in , /von

Miriam Fahimi (vom UZ D!ARC), Doktorandin im Projekt NoBIAS – AI without Bias (NoBIAS ), erzählt im Standard-Interview von den Gefahren von Künstlicher Intelligenz. Ihre Einsichten sind Teil ihrer laufenden Doktorarbeit zu Diskriminierung und Fairness in KI-Systemen (Visitenkarte ) im Fach Wissenschafts- und Technikforschung.

„Die Welt und unsere Gesellschaft sind von Ungleichheit geprägt, wie sollen da die Daten für eine künstliche Intelligenz völlig objektiv sein. Das ist utopisch“, erklärt Miriam Fahimi, die am Digital Age Research Center der Universität Klagenfurt zum Thema ‚faire Algorithmen‘ forscht. (Auszug aus dem Interview)

Den gesamten Artikel gibt es hier zu lesen: Zeitalter der KI 

Wir gratulieren Miriam Fahimi zu Ihrem Interview und freuen uns über weitere Erkenntnisse aus ihrer laufenden Arbeit.

D!ARC Lectures: Cryptographic Engineering Research: Navigating Responsibility Univ.-Prof.Dr. Elisabeth Oswald

/in , /von

12th January 2023    17:30 Uhr/ 5.30pm     Hörsaal 2/ HS 2

 

Cryptographic Engineering Research: Navigating Responsibility

Univ.-Prof. Dr. Elisabeth Oswald

 

Abstract

This talk is about challenges that arise when engineering systems in such a way that as little information as possible is leaked about cryptographic secrets. Over the years a range of mathematical and engineering techniques have been researched (and in part deployed) to account for, and mitigate, information leakage. Research in this area requires to carefully consider how developed techniques (that describe and analyse information leakage) not only help developers and evaluators, but if and how these can play into the hands of potential adversaries.

 

CV

Elisabeth Oswald completed her PhD in Technical Mathematics at the Technical University in Graz. Thereafter she took up a lecturing position in the Computer Science Department at the University of Bristol, where she established a research group in the area of applied cryptography, with an emphasis on analysing side channels. Eventually she was promoted to the first female chair in the Bristol Computer Science department. Her scientific accomplishments were honoured by an EPSRC Leadership Fellowship, an ERC Consolidator grant, and a number of best paper awards. She serves as associate editor of the two most influential journals in the are of cryptography, and participates regularly in leading functions for research funding institutions. Since 2019 she holds a chair in Cybersecurity research at the University of Klagenfurt.

 

For those who can only participate in this D!ARC Lecture online, see added the corresponding link for the live stream:

Für jene, denen nur eine online-Teilnahme an dieser D!ARC Lecture möglich ist, finden Sie anbei den entsprechenden Link für den Livestream:

https://classroom.aau.at/b/sag-893-nqz-qhq

D!ARC Lectures 24.11.2022 Automating Criminal Justice

/in , , /von

24th of November 2022     4 till 6 p.m.   HS 5

Abstract

Prof. Dr. Aleš Završnik

Automated decision-making processes already influence how decisions are made in the financial industry, as well as in education and employment. Applied to social platforms, they have contributed to the distortion of democratic processes, such as general elections. This trend is a part of “algorithmic governmentality” (Rouvroy and Berns, 2013) and the increased influence of mathematics on all spheres of our lives. It is a part of “solutionism”, whereby tech companies offer technical solutions to all social problems, including crime. Despite the strong influence of mathematics and statistical modelling on all spheres of life, the question of “what, then, do we talk about when we talk about ‘governing algorithms’?” (Barocas et al., 2013) remains largely unanswered in the criminal justice domain. How does the justice sector reflect the trend of the “algorithmisation” of society and what are the risks and perils of such? The purpose of the talk is to, first, examine the more fundamental changes in knowledge production in criminal justice settings occurring due to over-reliance on the new epistemological transition, and second, to show why automated predictive decision-making tools are often at variance with fundamental liberties and also with the established legal doctrines and concepts of criminal procedure law.

CV

Dr. Aleš Završnik is the Director of the Institute of Criminology at the Faculty of Law in Ljubljana and Full Professor at the Faculty of Law University of Ljubljana, Slovenia. His research interest lay in the intersection of law, crime, technology, and fundamental rights.

Dr. Aleš Završnik was EURIAS / Marie Curie Fellow at the Collegium Helveticum (ETHZ) in Zürich (2017/18) and also postdoctoral fellow of the Norwegian Research Council at the University of Oslo and at the Max-Planck-Institute für ausländisches und internationals Strafrecht in Freiburg i. Br.

Among several others, he led a research project “Automated Justice: Social, Ethical and Legal Implications” (Slovenian Research Agency, 2018–21) and “Human Rights and Regulation of Trustworthy Artificial Intelligence” (Slovenian Research Agency and several Ministries, 2019–21).

He edited several books, such as Big Data, Crime and Social Control (Routledge, 2018) and Automating crime prevention, surveillance, and military operations (Springer, 2021) and published several articles in the field of law, technology and human rights, such as the article “Algorithmic Justice” (European Journal of Criminology, 2019). He organised several conferences in these research areas over the last 15 years, e.g. Automated Justice: Algorithms, Big Data and Criminal Justice Systems in Zürich (2018) and Big Data: Challenges for Law and Ethics in Ljubljana (2017).

Završnik collaborated with the European Commission for the Efficiency of Justice (CEPEJ) on the preparation of the “Ethical Charter on the Use of Artificial Intelligence in Judicial Systems” (2018). He is an independent Ethics Expert with the European Research Council (ERC) (from 2017).

Address:

Prof. Dr. Aleš Završnik

Institute of Criminology at the Faculty of Law, Ljubljana

Poljanski nasip 2

SI-1000 Ljubljana, SLOVENIA

P: +386 1 4203 251

E: ales [dot] zavrsnik [at] pf [dot] uni-lj [dot] si

F: +386 1 4203 245

W: www.inst-krim.si

„Workshop: Ethnomethodology in the digital world with Andrei Korbut”

/in , , /von

Das D!ARC freut sich, den bevorstehenden Workshop über ethnomethodologische Ansätze zur Erschließung digitaler Welten ankündigen zu können. Der Workshop wird von Andrei Korbut, PhD, geleitet, der assoziierter Forscher in der Gruppe Humanwissenschaft des Digitalen ist.

Der Workshop ist offen für alle akad. Stufen. Für die Teilnahme sind keine Vorkenntnisse in Ethnomethodologie erforderlich. Der Workshop wird auf englisch abgehalten. Bei offenen Fragen oder für die Anmeldung am Workshop senden Sie bitte eine E-Mail an: clara [dot] hoestermann [at] aau [dot] at

Wir freuen uns darauf, Sie bei der Veranstaltung begrüßen zu dürfen.

Der Workshop findet statt: 15. November 2022 von 15-17 Uhr im Raum N.2.57 (AAU Campus)

Weitere Information finden Sie hier.The Ethnomethodology of digital worlds

 

Andrei Korbut (Assoc. Researcher, D!ARC)

is an independent researcher with primary interest in ethnomethodological studies of human–computer interaction and, in particular, user encounters with artificial intelligence. He defended his PhD in Sociology in the Higher School of Economics (Russia) in 2014, and then worked as a researcher in several academic institutions in Russia. He applies ethnomethodology and conversation analysis to the research in digitalization of labor, communication with voice interfaces, and social transformation of homes by smart technologies.