07/08/2026
Somewhere in your business, an AI agent just made a decision. No one told it to do that specific thing in that moment; it decided on its own based on how it was built.
New Australian research puts a number on what most technology leaders have quietly suspected: most businesses can't explain to a regulator what their AI agents did, or why. Nearly one in five can't say who'd be held accountable if it went wrong.
Regulators have already flagged the gap. Insurers are paying closer attention to AI-related exposures when assessing risk inside policy wording that most businesses haven't reread since their agents went live.
The real question isn't whether to keep running agentic AI. It's whether anyone's checked which policy would actually respond if one of those agents got it wrong.
We break down what the data shows, and how we help leadership teams turn assumption into evidence before a claim forces the conversation.
Read the full blog here: https://www.knightcorp.insure/insights/your-ai-agent-just-made-a-call-no-one-signed-off-on-it