An AI system is set to operate in more than one jurisdiction. What should the team do to match good practice?
Select an answer to reveal the explanation.
Short Explanation
Different jurisdictions genuinely disagree, so you can't pick one policy and hope it covers everything. Map your governance to each jurisdiction's actual laws and standards.
Full Explanation
The correct approach is to map governance activities to the specific laws and standards of each relevant jurisdiction, rather than assuming one policy fits all of them. Legal and regulatory requirements differ by jurisdiction, so a system operating across several needs its governance mapped to each one's actual requirements, not to a single assumed policy.
Applying the policy of the jurisdiction where the company is headquartered to every deployment is incorrect because headquarters-jurisdiction rules do not automatically satisfy the requirements of every jurisdiction where the system actually operates.
Applying the least restrictive applicable jurisdiction's rules everywhere, to simplify compliance, is incorrect because applying the least restrictive rule everywhere leaves the organisation out of compliance wherever a stricter jurisdiction's requirements were not met.
Deferring jurisdictional mapping until a regulator raises a specific concern is incorrect because waiting for a regulator to raise a concern means the gap existed, undetected, the whole time beforehand.
For the exam, map governance to each applicable jurisdiction rather than assuming a single policy generalises.