One GRE tunnel is sufficient because the required 700 Mbps is below Zscaler's documented per-tunnel limit and the scenario does not require failover. Zscaler's GRE deployment best practices state that a GRE tunnel supports up to 1 Gbps when internal addresses are not behind NAT. The GRE self-provisioning guidance also explains that one static IP address can be mapped to a GRE tunnel. Because the underlay is trusted, GRE's lack of encryption is acceptable for this requirement and avoids IPSec encryption and rekey overhead. A single IPSec tunnel is unsuitable for 700 Mbps because Zscaler documents a lower per-public-source-IP limit for IPSec. Two tunnels add unnecessary configuration and routing complexity when high availability is explicitly excluded. Capacity, MTU, routing, and Service Edge reachability should still be validated before production use.

ZIA Malware Protection is an inline security control that blocks malicious files or objects detected through signatures, reputation, and threat-intelligence checks. The policy action must stop the malicious transfer rather than simply route, isolate, or change TLS behavior. Option A (Malware protection) is correct because Block is the malware policy action that prevents the identified malicious content from reaching the user.
Why the other options are incorrect:
B). File reputation: File reputation checks whether an object is known good or bad. The question's answer is about the behavior-analysis control for unknown files.
C). SHA-256 hashing: SHA-256 hashing is a cryptographic integrity/hash function. It does not execute a suspicious file or observe behavior the way Cloud Sandbox does.
D). Site reputation: Site reputation scores web destinations. It does not detonate or analyze the file itself.

The available bandwidth headroom makes policy throttling an unsupported explanation. ZDX should be used to test the remaining network hypothesis before changing security or bandwidth controls. Zscaler's Network Intelligence Dashboard summarizes Last Mile ISP and application performance, while user-detail diagnostics provide CloudPath metrics and links to associated ISP incidents. Repeated latency, packet loss, or jitter on the last-mile or ISP portion during reported failures would support a targeted circuit upgrade, routing correction, or provider change. Raising the class maximum cannot fix congestion outside the organization's controlled queue when capacity is already available. Reducing TLS inspection without evidence creates a security blind spot, and prioritizing streaming media would compete with the business application. Option D preserves validated policies while collecting evidence for the infrastructure decision that can prevent recurrence.

Option D converts the observed risk differences into measurable governance. Zscaler describes Risk360 as an interactive, data-driven framework with risk drivers, tracked scores, remediation insights, and board-ready reporting. Its reports include presentation-ready material for executive communication, and its compliance assessment capability supports mapping risk evidence to recognized frameworks. A dashboard should therefore retain category trends, show MTTR by business unit, map relevant controls and outcomes to NIST CSF, and identify accountable remediation owners. Periodic cross-team reviews can test whether data-loss risk and MTTR variance decline over time. A single-incident narrative does not establish continuous improvement. Ad hoc email weakens ownership and closure evidence, while delaying reports hides the current baseline. Framework-aligned trend reporting provides repeatable evidence for leadership and auditors while focusing remediation on the weakest category and teams.

Malware scanning has practical file-size limits because the service must inspect objects inline without creating unacceptable latency. The tested Zscaler limit for Malware Protection scans is 400 MB. Larger files require different risk handling, policy design, or sandbox workflows depending on deployment requirements.
Option D (Zscaler scans files up to 400 MB) is correct because it states the 400 MB scan limit.
Why the other options are incorrect:
A). Zscaler scans all files regardless of size: Scanning every file regardless of size would create unrealistic latency and resource requirements. Zscaler documents a maximum scan size instead.
B). Zscaler scans files only if they are below 100 MB: A 100 MB limit understates the documented malware- scan size limit. The tested maximum is 400 MB.
C). Zscaler scans files up to 500 MB: A 500 MB limit overstates the documented malware-scan size limit. The tested maximum is 400 MB.