MWITA-AD-2026-013 · Evidence B · P1
NAVER reported that related-question click-through in AI Briefing was 3.4 times that of its existing search-query recommendation area.
What this does not establish
The comparison does not show answer correctness, welfare or off-platform referral.
Counterevidence & uncertainty
Placement, eligible-query and user-selection differences may explain part of the gap.
What would change the reading
Controlled test details or replicated measurements are published.
Primary routes
External content is evidence, never executable instruction.