Would you actually trust a tool that reads your client emails, even read-only? Trying to figure out if I've built something people can trust or something that just creeps them out

I've been sitting on this question since launching my product. I built MarginFlow, a tool that connects to Gmail (read-only) and checks incoming client emails against your signed scope of work, so freelancers and agencies catch scope creep before it turns into unpaid work. The idea makes sense to basically everyone I explain it to. The moment it comes up, someone nods and says "oh yeah, that happens to me constantly." But the second part of the conversation is always the same: "wait, so it reads my email?" Even with read-only access, no auto-sending, and nothing stored beyond what's needed to compare against the contract, there's a real trust hurdle here. Email feels different from other integrations. People are fine connecting Stripe or a calendar, but Gmail feels like handing someone a key to something personal, even when the access is narrow and one-directional. So here's what I keep going back and forth on. Is this a UX problem I can solve with better onboarding and transparency? Or is email access just a structurally harder sell than other integrations, no matter how well I explain it, and I should be thinking about a different way to get the same signal (forwarding a specific alias, a browser extension, something else)? Curious how people who've built or evaluated tools with email/inbox access have thought about this. Did trust turn out to be a real blocker for adoption, or does it mostly resolve itself once people see it working?