WorkBuddy's contract comparison feature: I trust it, but only halfway

I saw the release of the 2026 International AI Safety Report last week. Over a hundred experts listed emerging risks brought about by the rapid improvement in general AI capabilities. The report specifically noted that policymakers often lack sufficient scientific evidence to weigh the pros and cons. As I read it, I thought this dilemma applies just as much to everyday office tools. How many of us frontline users actually have enough evidence to judge whether an AI tool is reliable or not?