Where AI changes what's possible — and what's governable.
A monthly thread. Essays, scenarios, and exposure audits on the decisions leaders will have to make as capability outruns policy.
Start with an Exposure Audit
A 2-week review of where AI quietly erodes your current controls — and the three changes that matter most.
Latest in Governance Horizons
Five Governance Assumptions That Break When LLMs Learn From Your Data
Enterprise AI governance rests on a silent architectural assumption: that the model you reviewed at time T is the model running at time T+6 months. That assumption is now routinely false, and five pillars of assurance are collapsing quietly beneath organisations that haven't noticed yet.
10 min read readThe Three Things People Mean When They Say 'AI Memory' — and Only One Is a Governance Bomb
The industry talks about 'AI memory' as if it's one thing. It's three. Conflating them is how governance programmes sign off on risks they never identified — and how enterprises quietly accumulate GDPR exposure they cannot delete.
10 min read readGet Governance Horizons monthly
One essay a month. No other mail. Unsubscribe any time.