AI leaders navigating internal pressure to approve a steady stream of proposed initiatives are increasingly using a specific rhetorical and decision-making move: rather than an outright rejection, responding with a defined "not yet" - a clear statement of what specific condition would need to change before the proposal gets reconsidered, with a set date to revisit it.
Why outright rejection is politically costly
Flatly rejecting a colleague's proposed AI initiative can create lasting friction and discourage future proposals from that person or team, even when the rejection was well-justified. This is a real organizational cost that a purely rational evaluation of the proposal's merits doesn't capture.
Why a specific 'not yet' works better
A clearly conditioned deferral preserves the relationship and keeps the door genuinely open, while still preventing an unready initiative from consuming resources before its time. Crucially, it requires actual specificity - a vague "let's revisit this later" functions as a soft no without providing any of the political benefit of a real conditional deferral, since the proposer has no clear sense of what would actually change the outcome.
What makes this move genuinely useful rather than a stalling tactic
The condition has to be real, specific, and actually checked when the revisit date arrives - a vague deferral used simply to avoid the discomfort of a direct no is a different, less honest tactic that erodes trust once colleagues notice the pattern of deferrals never actually converting into approvals.
Why this skill matters for AI leadership specifically
The current environment generates an unusually high volume of AI proposals across most organizations, more than can reasonably be resourced simultaneously. Leaders who can defer well, honestly and specifically, are managing that volume more sustainably than leaders forced into a binary approve-or-reject decision on every one.