December 20, 2023

This Week in Responsible AI, Dec 20, 2023

This Week in Responsible AI, Dec 20, 2023

Tools

• Open AI's "Preparedness Framework"

• Risky Analysis: Assessing and Improving AI Governance Tools

• VonGoom: A Novel Approach for Data Poisoning in Large Language Models.

Data

• DICES Dataset: Diversity in Conversational AI Evaluation for Safety

• Ethical Considerations for Responsible Data Curation

• Largest Dataset Powering AI Images Removed After Discovery of Child Sexual Abuse Material

• The “Trolley Problem” Doesn’t Work for Self-Driving Cars

• China Using AI to Create Anti-American Memes Capitalizing on Israel-Palestine, Researchers Find

Fakes

• Facebook Is Being Overrun With Stolen, AI-Generated Images That People Think Are Real

• Take It to the Spank Bank

Policy / Law

• Human rights protections…with exceptions: what’s (not) in the EU’s AI Act deal

• Five things you need to know about the EU’s new AI Act

• Rite Aid Banned from Using AI Facial Recognition After FTC Says Retailer Deployed Technology without Reasonable Safeguards

• Google takes steps to prevent abuse of its generative AI tools during U.S. elections

• FTC Staff Report Details Key Takeaways from AI and Creative Fields Panel Discussion

• AI cannot patent inventions, UK Supreme Court confirms

• Model legislation for online civil rights

Privacy

• Marketing Company Claims That It Actually Is Listening to Your Phone and Smart Speakers to Target Ads

Fairness

• Repairing Regressors for Fair Binary Classification at Any Decision Threshold

Other

• Speculative F(r)iction in Generative AI

• These six questions will dictate the future of generative AI

Compiled by Leif Hancox-Li