OWAI INSIGHTS
Practical notes on what existing and new rules mean when an AI agent can take an action — and what an organisation should be able to show for it.
General information, not legal advice. Each note shows when it was last reviewed.
NOTES
EUROPEAN UNION / AI ACT · REVIEWED 28 SEPTEMBER 2026
What Articles 12, 14 and 26 of the EU AI Act mean when an AI agent can act, the dates after the 2026 Digital Omnibus, and how to evidence oversight in practice.
UNITED KINGDOM / FINANCIAL SERVICES · REVIEWED 28 SEPTEMBER 2026
The FCA regulates AI through existing rules. What the Consumer Duty and the Senior Managers and Certification Regime mean when AI agents act for a regulated firm.
FROM OUR RESEARCH
PREPRINT
A stopping point is not enough. The party at it needs presently valid, traceable authority, and the stop must actually change what can happen.
PREPRINT
Governance of AI-enabled operational systems has to live in the architecture that decides which state transitions are admissible — including across sequences of individually acceptable steps.
TECHNICAL NOTE
Inference may propose actions but never authorise them. Authority is resolved outside the model, and execution is impossible without it.
