When seeking help or advice, one naturally wishes to get that aid from someone who does the task well, from an expert. Determining whether someone merits that label is not trivial, and the judgment can have important consequences. Experts in most domains, from athletes to plumbers, need to exhibit a high level of performance to maintain their status. But there is another group of professionals that is never examined. The expert status of such experts is conferred via criteria such as education or experience. While their credentials may be challenged, their work itself is not seen to bear upon their status. We refer to this group as privileged experts. These experts make assessments or predictions whose correctness is rarely evaluated. Even if prophecy fails, errors are forgiven. We account for this non-evidentiary perspective by positing a widely-held misconception that expertise generalizes. We outline a task-specific, performance-based, limited-term certification scheme that can supplant privilege.