Topics

Browse posts by subject.

• The agentic state 11

  Software acting for citizens or public agencies, including questions of delegation, control and responsibility.

• State capacity 48

  Whether governments have the staff, systems and authority to use AI effectively.

• Service delivery 29

  AI in benefits, permits, health care and other public services.

• Democracy and participation 23

  Evidence on deliberation, citizen engagement and whether participation changes public decisions.

• Development 51

  The costs and benefits of AI in countries that mostly buy the technology rather than build it.

• Regulation 34

  The laws, liability regimes and enforcement tools governments use for AI.

• Procurement 17

  Contracts, purchasing rules and vendor relationships in public-sector AI.

• Agents 30

  What software agents can do, where they fail and who is responsible for their actions.

• Evaluation 27

  Tests that distinguish a demonstrated result from a claim about AI.

• Compute 16

  Access to chips and computing capacity, including who controls the infrastructure.