Consent Must Travel With the Data: A Rule for AI Personalisation
Responsible AI personalisation requires consent, purpose and revocation to travel with data as it moves from collection to inference and action.
Category: AI Ethics
OpenAI’s Misalignment Reporting Framework Signals a New Phase for AI Safety
OpenAI has introduced a framework for investigating and reporting model misalignment, raising an important question about how AI incidents should be documented.
Do Not Give an AI Agent the Keys to Everything: Build a Permission Budget
An AI agent needs enough access to finish its task—not your whole digital life. Build a permission budget with limits, approvals, logs and rapid revocation.
Small Business AI Governance: A Simple Policy That Works
Create a practical AI policy for your small business covering privacy, accuracy, human review, approved tools and accountability.
Why Do Some People Feel More Heard by AI? New UK–China Research Examines the Psychology of Listening
A new cultural case study from Cambridge and SOAS examines how a Chinese migrant woman living across China and the UK used generative AI for emotional support. The paper argues that AI may sometimes function as a temporary “care-patch” where human listening feels scarce, delayed or burdensome.
The AI Mirror: Effects Analysis Framework, AI Regulation and the Human Signals Behind the Machine
Mary Oge Chuks introduces the Effects Analysis Framework (EAF), a psychological and systems-based method for analysing AI through signals, observable effects, framing shifts and institutional outcomes — and asks why humanity regulates the reflection without examining the signals it feeds into the mirror.
Give Your AI Agent a Budget for Decisions
Set clear limits for AI agents: define what they may decide, when they must stop and how people can review the actions they take.
A Confident Answer Still Needs a Place for Disagreement
Keep room for disagreement in AI-assisted decisions with a simple record of alternatives, missing evidence and the reasons behind your final choice.
The Public-Service AI Test: Can People Complete the Ordinary Task?
A policy proposal for judging public-service AI through complete user journeys, including people who need a different way to access help.
AI Infrastructure Has a Neighbourhood: Ask Who Shares the Benefits
AI infrastructure decisions reach beyond server buildings. Communities need clear questions about local services, resources and promised benefits.