Enshittification short book review from a British perspective
A workshop on the 'AI Amplifier Effect' paper showed how AI relationships can reinforce either recovery and confidence or a negative spiral, depending on the user's starting state of mind.
A 17.3% hike to a basic M365 licence, folded in under a repriced Copilot/Cowork model, looks like the start of Microsoft's enshittification playbook.
Connecting Anthropic's J-space research to anthropomorphism and 'Seemingly Conscious AI' risk, this post asks whether employers already have a duty of care for AI-interaction mental health harm.
Oxford Internet Institute research found AI tools subtly shift the meaning of human-written drafts on sensitive topics, and no regulation currently addresses this kind of bias.
Reading 'Enshittification' alongside Microsoft's licence hikes and Anthropic's unsustainable running costs, this post argues enshittification is a matter of when, not if.
The concept of Minimum Viable Governance: the smallest amount of structure that lets your organisation move forward with AI Responsibly. A practical starting point for organisations who feel overwhelmed.
People want authenticity and transparency, but research shows that when companies disclose AI use, trust actually decreases. How do leaders navigate this paradox?
Generative AI is shaping more than productivity — it could quietly erode how our people think, learn, and work together. How do you respond as a business leader?
AI has become ever-present, but the question your employees, partners and customers are really asking is much simpler: can we trust you with AI?
Underfold