KTER Research · The Accountability series
Research
What an enterprise needs before it can stand behind an AI agent’s work. Five papers, one argument.
The series
Five gaps, one structure
Each paper names one way an enterprise ends up carrying risk it cannot see. Read together they describe a single structure that is missing.
AccountabilityWho answers for the output.
Blast radiusWhat the agent can reach.
ConsequenceWhat the agent would never do.
VisibilityWhether anyone understands it.
EvidenceWhether the claim of success can be relied on.
All papers
Paper 1 · 16 July 2026The Accountability GapWhy every enterprise AI deployment has a hidden liability problem.
Paper 2 · 23 July 2026Trust Has a Blast RadiusWhy boundaries have to exist before an agent runs, not after.
Paper 3 · 27 July 2026The Consequence GapWhy an agent can’t restrain itself the way a person does.
Paper 4 · August 2026The Visibility DebtWhy running an agent and understanding an agent are not the same thing.
Paper 5 · October 2026The Evidence GapWhat an enterprise needs before it can accept an agent’s claim of success.