How AI is breaking the British state: policy-levers lab

“To make the state less vulnerable requires tightening rules, closing loopholes and creating disincentives to AI-enabled excess. Byzantine appeals systems should give way to simpler ones. Money should be part of the solution.”

Drag a remedy onto its station, or tap a remedy then tap its station.

Drag a red stream past its defence — or use the levers above — then read the consequence below.

Excess events per week 1000

Legitimate-case clearance % 62%

Appeals backlog, weeks 8.2 wks

Public-trust index 35

Status quo: Byzantine appeals and weak enforcement leave the state overwhelmed — 1,000 excess events a week, 62% clearance, an 8.2-week backlog and public trust at 35.

How the model works: weekly flows are illustrative volumes derived from the captured post. “Tighten rules”, “Close loopholes”, “Simplify appeals” and “Money disincentives” each change a flow, the clearance rate, the backlog or funding. The trust index is a deliberately simple comparable, not an official measure.

“To make the state less vulnerable requires tightening rules, closing loopholes and creating disincentives to AI-enabled excess. Byzantine appeals systems should give way to simpler ones. Money should be part of the solution.” — The Economist