
Brent Mittelstadt @bmittelstadt.bsky.social
@b_mittelstadt
Professor of Data Ethics and Policy & Director of Research at @oiioxford @uniofoxford. Founder of Governance of Emerging Technologies programme @OxGovTech.
ID: 799390015
https://www.oii.ox.ac.uk/research/governance-of-emerging-technologies/ 03-09-2012 01:31:52
943 Tweet
6,6K Followers
287 Following

There's a really interesting paper published today that questions whether AI chatbots should have a legal duty to tell the truth. For New Scientist I asked the authors - and experts - how practical that is newscientist.com/article/244286ā¦

News alert! Large Language Models pose a risk to society and need tighter regulation, say Oxford researchers Sandra Wachter [email protected] Brent Mittelstadt @bmittelstadt.bsky.social Chris Russell Governance of Emerging Technologies Oxford Internet Institute OxfordSocialSciences University of Oxford. oii.ox.ac.uk/news-events/do⦠1/4


Delighted to see coverage of our new paper on truth and LLMs in New Scientist ! newscientist.com/article/244286⦠Oxford Internet Institute University of Oxford Governance of Emerging Technologies Sandra Wachter [email protected] Chris Russell

We are thrilled to announce that Brent Mittelstadt Brent Mittelstadt @bmittelstadt.bsky.social, Associate Professor and Director of Research at the Oxford Internet Institute, University of Oxford, will be speaking at #ODSCEurope this fall! ā Register Now ā Limited Passes Availableš„ hubs.li/Q02KdLty0


Such an honour to be featured in this nature Vivien Marxšš¬ā ļø article w/Brent Mittelstadt @bmittelstadt.bsky.social Chris Russell on our work on GenAI, truth & hallucinations

Do large language models have a legal duty to tell the truth? This new #RSOS paper investigates: ow.ly/uasi50SVZq0 #LLMs #AI Brent Mittelstadt @bmittelstadt.bsky.social Sandra Wachter [email protected] Chris Russell Wendy Hall


I have a new paper out "Limitations and Loopholes in the EU AI Act and AI Liability Directives: What This Means for the European Union, the United States, and Beyond" in The Information Society Project @yalej Yale Journal of Law & Technology Yale University law.yale.edu/sites/default/ā¦

I am delighted to share that I have been awarded the title of Full Professor by University of Oxford Oxford Internet Institute! I am now Professor of Data Ethics and Policy. Thank you to everyone that made this possible, colleagues, friends, family, teachers, and especially Sandra Wachter [email protected] & Chris Russell

Do large language models have a legal duty to tell the truth? In our next behind the paper seminar on 16 Oct at 2pm (Europe/UK), authors Sandra Wachter [email protected] and Brent Mittelstadt @bmittelstadt.bsky.social speak with Mirco Musolesi about their recent article. Find out more & sign up: cassyni.com/events/VFAWmocā¦

Brent Mittelstadt @bmittelstadt.bsky.social Sandra Wachter [email protected] Chris Russell Wendy Hall Join us on Wednesday 16 October at 2pm (Europe/UK) to hear authors Sandra Wachter [email protected] and Brent Mittelstadt @bmittelstadt.bsky.social discuss this article with #RSOS editor Mirco Musolesi - find out more & sign up here: cassyni.com/events/VFAWmoc⦠#LLMs #AI

Join me & Brent Mittelstadt @bmittelstadt.bsky.social at the "behind the paper" virtual seminar The Royal Society Royal Society Publishing today 16 October at 14.00. We will present our latest work "do large language models have a legal duty to tell the truth?"cassyni.com/events/VFAWmoc⦠Oxford Internet Institute Berkman Klein Center for Internet & Society HIIG Berlin


š Paper news! Excited that our paper, with Brent Mittelstadt @bmittelstadt.bsky.social & Chris Russell was accepted at the NeurIPS Safe GenAI workshop! In this work we explore how fine-tuning can impact toxicity rates in language models... š§µ


Prime example of what I Brent Mittelstadt @bmittelstadt.bsky.social Chris Russell termed careless speech. Developers must be responsible 4 outputs of their systems. False AI outputs can poison science & public policy tinyurl.com/48kpbm9d see Do LLMs have a legal duty to tell the truth?tinyurl.com/3dae48yx

#HealthcareAI developers, help shape trustworthy AI! Take 15 mins to share insights with researchers from the Oxford Internet Institute and Uni of Reading in a study supported by NHSx, Wellcome Trust & Sloan Foundation. Enter a raffle to win Ā£250! šoii.qualtrics.com/jfe/form/SV_2gā¦

Another example of what I Brent Mittelstadt @bmittelstadt.bsky.social Chris Russell termed careless speech. Subtle hallucinations are dangerous & developers are not (yet) liable for them. We argue they should. See paper: Do LLMs have a legal duty to tell the truth? tinyurl.com/435jba5p tinyurl.com/4wpske2s


NEW: OII Podcast Episode 5 with Sandra Wachter [email protected] + Brent Mittelstadt @bmittelstadt.bsky.social discussing the current state of AI regulation, potential negative consequences of AI proliferation, and what needs to change about how AI is governed. Listen here: podcasts.ox.ac.uk/how-ai-being-rā¦
