Insuring AI: Not “How Much” — “Which Policy”
The hardest question in insuring AI isn’t “how much?” It’s “which policy?” Risk is migrating from physical to digital, and digital risks refuse to stay in their lanes. An AI failure event — is it cyber? Tech E&O? Physical damage downstream? The honest answer is often all of the above, overlapping. That’s a real problem, because the one thing every client wants from an insurance contract is clarity. Lucy Pilko’s prediction rhymes with history: cyber insurance evolved out of silent cover that already existed in policies — AI cover will evolve the same way, out of today’s wordings into new policy language. The talent implication: underwriters who think fluidly across once-clean niches are about to be the most valuable people in the building.