AI governance
6 posts
Anthropic, OpenAI, and DeepMind grade their own models' danger
AI labs publishing dangerous-capability evals turns safety disclosure into marketing, and the self-graded scorecards drift toward a race to the bottom.
Unsealed court briefs turned training data into evidence
Unsealed briefs in the authors' case against Microsoft and OpenAI reveal training-data provenance as a board-level liability surface, not a technical detail.
Resignations are signals, not scandals
How to read a high-profile resignation from an AI safety lab like Anthropic - what it signals, what to ignore, and what to watch in the months after.
Alibaba moves to ban Claude Code over backdoor risk
A reported ban on a widely-used AI development tool reframes third-party AI access as a board-level exposure most organizations have not measured.
Your AI features are now your attack surface
Meta has confirmed over 1,000 Instagram accounts were compromised through abuse of its AI chatbot - a board-level view of the control failure.
You still own every decision you automated.
When automation decides in threat detection and response, the judgment moves but the accountability does not - and the organisation owns every outcome it cannot explain.