A stormwater bureau must keep inference in an approved region for data-residency rules but still wants managed capacity. Which option best fits?
Select an answer to reveal the explanation.
Short Explanation
You can keep managed capacity and meet the residency rule by choosing a sovereign or residency-compliant cloud region. That lets you avoid unrestricted global endpoints without abandoning generative AI altogether.
Full Explanation
Cloud hosting in a sovereign or residency-compliant region best fits the bureau’s need to keep inference in an approved region while retaining managed capacity. Stormwater data-residency rules constrain where prompts and outputs may be processed; compliant regional or sovereign cloud offerings can meet those rules while providing managed capacity. A managed public endpoint in any available region, without residency restrictions or controls on request location, fails the policy constraint and may process civic data in a disallowed jurisdiction. Shadow IT use of personal accounts in foreign regions bypasses approved controls, billing, and audit trails. Full abandonment of generative AI regardless of controls is unnecessary when residency-compliant cloud hosting is available. Exam caveat: a residency requirement does not automatically rule out cloud hosting; assess whether the region and service meet the stated rules while preserving managed capacity. Practical check: pin inference to approved regions in vendor configuration, verify that subprocessors and failover remain within the approved geography, update DPAs for residency obligations, block noncompliant endpoints at the network layer, and retain evidence that inference stayed in-region during normal and failover tests.