Liability and insurance
Someone answers when an AI system gets it wrong.
capabilityEmerging Open on the map →
Facts
- Lane
- Trust and governance
- Wave
- 9 of 12 (build order, not a date)
- Needs
- 17 boxes all the way back through hard links
- Unlocks
- 0 boxes downstream through hard links
- Leverage
- None
- Cluster
- Trust plumbing and the agent economy
Evidence as of October 2026
A Singapore IMDA paper in May 2026 examined which of developer, provider or deployer answers when an agent acts in unintended ways, and recorded disagreement. Gervais and Nay argue liability should trace to the humans behind the system through existing agency law.
Detail boxes 1
EU strict liability for AI software
The revised EU Product Liability Directive makes software and AI systems products under no-fault liability, with court-ordered disclosure and presumptions of defect.
Directive 2024/2853 must be transposed by 9 Dec 2026; a defect is presumed if a defendant fails to disclose evidence, and open source supplied outside commercial activity is exempt.
Around it
Loops and off-ramps
Openings researchers' judgement, October 2026 scouting
Startup room 57/100 · rank 20 of 124 Research room 48/100 · rank 69 of 124
- Depth
- light
- Players
- AIUC ($40m A, AIUC-1); Armilla ($25m); Lloyd's and reinsurers
- Wedge
- Actuarial pricing of correlated AI failures via the MGA-plus-certification template AIUC and Armilla use. startup