January 12, 2024

This Week in Responsible AI: Jan 12, 2023

This Week in Responsible AI: Jan 12, 2023

Art

• 'Magic: The Gathering' Publisher Denies, Then Admits, Using AI Art In Promo Image

• Artists are making creative companies apologize for using AI

• Work inside the machine: How pre-saves and algorithmic marketing turn musicians into influencers

Security

• NIST Identifies Types of Cyberattacks That Manipulate Behavior of AI Systems

• This AI Chatbot is Trained to Jailbreak Other Chatbots

Law/Policy

• America’s Next “Stop Model!”: Model Deletion

• Experiments with Midjourney and DALL-E 3 show a copyright minefield

• Election officials worry about the potential use of AI to spread misinformation

• OPENAI QUIETLY DELETES BAN ON USING CHATGPT FOR “MILITARY AND WARFARE”

Privacy

• Privacy Issues in Large Language Models: A Survey

• AI Companies: Uphold Your Privacy and Confidentiality Commitments

• Face-off in Britain over controversial surveillance tech

Fabrications

• Hallucinating Law: Legal Mistakes with Large Language Models are Pervasive

• BEWARE OF BOTSHIT: HOW TO MANAGE THE EPISTEMIC RISKS OF GENERATIVE CHATBOTS

• Deepfaked Celebrity Ads Promoting Medicare Scams Run Rampant on YouTube

• YouTube is cracking down on AI-generated true crime deepfakes

• AI Is Coming for the Influencers

• Nikon, Sony and Canon fight AI fakes with new camera tech

• Deepfakes and Child Safety: A Survey and Analysis of 2023 Incidents and Responses

• Memento Mori: Making kin with AI at the end of the world

• Hackers Steal Dog Rescue's Facebook Page, Turn It Into AI Content Farm

• Content Credentials Will Fight Deepfakes in the 2024 Elections

Bias

• "we find substantial misalignment between the views reflected by current LMs and those of US demographic groups: on par with the Democrat-Republican divide on climate change."

• Assigning LLMs different social roles affects the accuracy of their answers

Education

• A Realist Review of Undergraduate Student Attitudes towards Ethical Interventions in Technical Computing Courses

• "Amid South Africa’s perfect storm of few universities, racial inequities and economic disparities, AI students find support in a not-so-grassroots grassroots movement."

Other

• Impossibility Theorems for Feature Attribution

• Chicago shouldn’t renew its ShotSpotter contract

• Algorithmic Risk Assessment in the Hands of Humans

• LAION-5B, Stable Diffusion 1.5, and the Original Sin of Generative AI

• Can LLMs make medicine safer?

• "Twitter’s engagement-based ranking algorithm amplifies emotionally charged, out-group hostile content that users say makes them feel worse about their political out-group."

• Impossible AI

• Advancing transparency: Updates on responsible AI research

• Measuring Justice: Field Notes on Algorithmic Impact Assessments

• The Algorithms Too Few People Are Talking About

Compiled by Leif Hancox-Li