How can we determine whether GenAI has identified all significant problems in a legacy system when some problems may be unknown even to the organization?
What happens when essential source programs, documentation, historical records, or experienced specialists are permanently unavailable?
How can we measure the completeness and reliability of GenAI's investigation when no fully verified reference exists?
Can GenAI distinguish missing knowledge from information that exists but has not yet been discovered?
How should conflicting evidence from programs, data files, documentation, and human experts be investigated and resolved?
What procedures should apply when GenAI identifies a potential problem but cannot produce sufficient evidence to establish its cause?
How can we ensure that unresolved problems, uncertainties, and rejected findings remain traceable within the baseline rather than disappearing from future investigations?
At what point should an organization decide that further investigation is no longer productive and refer unresolved matters to authorized human decision-makers?