UBC CS Public Lecture: Regulation in the Age of Cloud Computing & Generative AI

Keynote speakers and panelists:

Prof. Lilian Edwards, Professor of Law, Newcastle University

Prof. Christopher Millard, Professor of Privacy and International Law, University of London

Prof. Vered Shwartz, Assistant Professor, UBC Computer Science

Moderator: Prof. Thomas Pasquier, Assistant Professor, UBC Computer Science

RSVP: Please sign up below if you are interested in attending.  Refreshments will be served starting at 5 pm.

Description: Advancements in Machine Learning and Cloud Computing have paved the way for a global race to develop Generative AI systems. Our panel discussion will delve into its societal effects and implications for regulation. Prof. Lilian Edwards (Newcastle University) will kick off the discussion, bringing her expertise in Internet law, intellectual property, and artificial intelligence. Following her, Prof. Christopher Millard (Queen Mary University of London), an expert in cloud computing law, will explore the sovereignty implications of growing dependence on US hyperscale cloud providers. Joining them on the panel is Prof. Vered Shwartz from UBC Computer Science, an expert in Natural Language Processing currently working on developing culturally-aware generative AI. This public event, part of the graduate course on Accountable Computer Systems at UBC, is made possible through the support of a Killam Connection Award.

Event details:

Keynote 1:

Title: Cloud Sovereignty: A Transatlantic Perspective

Presenter: Prof. Christopher Millard

Abstract: Canadian concerns regarding digital sovereignty date back to the early 1970s, while EU data protection laws have a ‘long-arm’ territorial reach and contain complex restrictions on international data transfers. Recent disruptions to global supply chains due to regional conflicts and trade disputes have led to heightened awareness of dependencies on the ‘big three’ US cloud providers. This talk will examine the complex drivers for assertions of cloud sovereignty and assess whether calls for a ‘European cloud’ make strategic, regulatory, or commercial sense.

See here for Prof. Millard's biography.

Keynote 2:

Title: Fixing ChatGPT’s Bad Dreams: Remedies for generative AI and Reputational Harms

Presenter: Prof. Lilian Edwards

Abstract: A recent 2023 episode of the notorious Black Mirror series “Joan is Awful” suggested that we might find ourselves legally signing away the right to dramatize our own lives, via AI generated quantum magic,  as part of the terms of service of a streaming platform we subscribed to. While legally unlikely, the show illustrates vividly a current issue – the wholesale tendency of large language models to scrape in unsupervised fashion biographical data from the public Internet and then to generate frequently false, distorted and sometimes upsetting results. Generative tools can also be used deliberately (as in the Joan episode) to create unauthorized, abusive or even sexualized versions of real people.  This is affecting everyone from politicians to actors to most recently, Taylor Swift, and has potential to destabilise elections, harass women, pollute the public infosphere and scam the public.

What control do we have over these “hallucinated” versions of what we say, what we look like, who we are? How can we separate generated fiction from “real” fact? Do existing legal instruments help us with this kind of control over our public persona? Using a real documented example involving a colleague, in this paper I focus on whether current laws eg European data protection law, defamation, copyright and personality rights are adequate to deal with these kinds of issues going forwards and whether any global trends can be seen in a very unharmonized legal sphere. We propose that instead of worrying about existential risk, we should be thinking more prosaically about a harmonized global law of personality rights.

See here for Prof. Edwards' biography.

Panetlist: Prof. Vered Shwartz

Biography: Prof. Shwartz is an Assistant Professor of Computer Science at the University of British Columbia, and a CIFAR AI Chair at the Vector Institute. Her research interests focus on natural language processing, with the fundamental goal of building models capable of human-level understanding of natural language. She is interested in computational semantics and pragmatics, and commonsense reasoning. She is currently working on learning to uncover implicit meaning, which is abundant in human speech, developing machines with advanced reasoning skills, multimodal models, and culturally-aware NLP models.

Moderator: Prof. Thomas Pasquier

Biography: Prof. Pasquier is Assistant Professor in the Computer Science Department at the University of British Columbia. His research addresses a broad array of issues in computer systems and security. He is particularly interested in the challenges of building more trustworthy, observable and transparent systems. He works on topics such as auditing, intrusion detection, and systems optimization in various environments such as operating systems, cloud computing, and the internet of things.