Latest
Where the oversight went
Four out of five CIOs say they have lost oversight of the AI agents in their organisations. The fix is not more monitoring. It is a structure with a name on it.
The named reviewer already exists
Governance frameworks keep asking who owns the AI agent. The answer is not a new hire or a committee — it is the named, credentialed expert who already reviews the work.
One owner, one metric
Mid-market AI deployments stall when ownership is diffuse. The operating model that works assigns one owner, one decision-rights holder, and one metric.
Gate by consequence
The standard picture of safe AI is a human approving every action. That picture breaks at scale. On the shift from blanket approval to expert supervision, and where judgement actually belongs.
The underwriting gap
Insurance carriers began attaching generative-AI exclusions to commercial policies in January 2026. Companies deploying AI agents are discovering their coverage no longer responds. The oversight layer is becoming an insurance requirement.
The denial arrives at machine speed
Payers now decide prior authorisations with AI in seconds. Providers still answer with fax machines and phone queues. The asymmetry is the story — and the staffing question it creates.
Paying for the ramp
A senior hire's salary starts on day one. Their output does not. The 4–6 month ramp is the quiet line item in every hiring decision — and it is negotiable.
Between pilot and production: where mid-market AI stalls
Most AI pilots work. Few ship. Why judgement — not intelligence — is the scarce resource between pilot and production.