Risk briefings, research and draft standards for people responsible for how organisations use AI.
Operational risk briefing
FAA ground stop shows why backup systems need their own test
A communications outage halted US departures nationwide. CAIRMC’s briefing asks teams to test the backup while the primary service is offline and check whether both paths share a carrier or cloud region.
How AI agents and shared suppliers can spread a failure
An agent with permission to change supplier details and prepare payments could carry one error through both steps. Adrian Dunkley examines how permissions and dependence on shared suppliers can spread losses.
What the Caribbean AI Risk Index can tell you before a risk decision
Country scores can help you choose what to investigate. The Council explains the evidence behind the scores, what a display correction changed and which questions need local data.
How would banks and regulators test a Caribbean AI risk standard?
CAIRMC’s consultation analysis gives banks and regulators questions to put to the draft, including what evidence an institution could produce to show its controls work.
Voluntary guidance from the US National Institute of Standards and Technology.
Common questions
Has CARICOM adopted the Council’s draft standard?
The Council’s consultation draft states that it has yet to be formally adopted or endorsed by CARICOM or any member state.
How should a risk team use the Caribbean AI Risk Index?
Use it to choose questions for further investigation. Check the data and the conditions in your own organisation before you use an index score to make a decision.