May 2, 2024

This Week in Responsible AI: May 2, 2023

This Week in Responsible AI: May 2, 2023

Hype

• Generative AI is still a solution in search of a problem

• Meta’s ‘set it and forget it’ AI ad tools are misfiring and blowing through cash

• Tackling AI Hyping

Evaluations

• The PRISM Alignment Project: What Participatory, Representative and Individualised Human Feedback Reveals About the Subjective and Multicultural Alignment of Large Language Models

• Holistic Safety and Responsibility Evaluations of Advanced AI Models

Synthetic media

• Meta AI falsely claims lawmakers were accused of sexual harassment

• Ex-athletic director accused of framing principal with AI arrested at airport with gun

• Flood of AI-Generated Submissions ‘Final Straw’ for Small 22-Year-Old Publisher

• Ads on Facebook, Instagram for explicit 'AI girlfriends' prompt Meta crackdown

• Consensual deepfakes

• Tech Companies Promise to Try to Do Something About All the AI CSAM They’re Enabling

Law/Policy

• There’s an AI Lobbying Frenzy in Washington. Big Tech Is Dominating

• Our government shouldn’t use the public sector as a guinea pig for AI

• Italy introduces entirely automated public tenders

• CISA’s chief data officer: Bias in AI models won’t be the same for every agency

• NIST announces a generative AI evaluation program and an AI Risk Management Framework Playbook.

• Keywords of the Datafied State

• How to govern AI — even if it's hard to predict

• Considerations for AI Opt-Out

• Surviving the AI Summer

• Use of Entity Resolution in India: Shining a light on how new forms of automation can deny people access to welfare

• Rishi Sunak promised to make AI safe. Big Tech’s not playing ball.

Other

• Simple probes can catch sleeper agents

• Tesla’s Autopilot and Full Self-Driving linked to hundreds of crashes, dozens of deaths

• A short and sweet account of how the Cold War influenced our contemporary attitudes that elevate prediction over understanding.

• From Thin to Thick: Toward a Politics of Human-Compatible AI

• AI-Related Risks Test the Limits of Organizational Risk Management

• Microsoft bans US police departments from using enterprise AI tool for facial recognition

Compiled by Leif Hancox-Li