ICONIQ's 2026 State of Scaling report names a quiet policy shift happening inside a lot of growth-stage companies. When someone leaves, the role increasingly doesn't get filled. The report describes how "attrition is increasingly being used as an opportunity to avoid, delay, or down-level backfills," with no-backfill policies "explicitly tied to AI-driven productivity improvements." The logic: omeone quits, AI has made the team more productive, so why replace them? Bank the savings and move on.
Finance leaders, applying this reflexively becomes a costly mistake. Not every open role is equal; the experienced, the judgment-heavy positions are ones you can’t afford to lose as these departures are a capability hole that AI can't actually fill.
ICONIQ's report documents that 70%+ of employees across portfolio companies are daily AI users, and that this adoption is contributing to rising revenue per employee. When AI genuinely absorbs a chunk of what a role used to do, not backfilling can be a legitimate efficiency decision. The report also notes companies using hiring freezes "while they evaluate AI-driven productivity gains before approving incremental headcount," which is a reasonable, disciplined way to manage growth.
For repetitive, high-volume work, this logic holds. If AI now handles the bulk of a transactional role's output, replacing that person one-for-one may genuinely be unnecessary. The report's own finance examples show AI absorbing large shares of contract processing, reporting, and reconciliation work, the kind of tasks where a no-backfill decision can make sense.
But the same report that documents the productivity gains also shows their limit. It finds that functional headcount mix has stayed largely unchanged, concluding that "while AI is improving productivity, it has not yet meaningfully reshaped operating models." In plain terms: AI is making existing people more productive, but it hasn't replaced the underlying structure of who does what. That's means the no-backfill reflex can quietly hollow out roles that the operating model still depends on. When the departing person owned judgment, review, or exception-handling, AI doesn't fill that hole. It just leaves it open.
Consider what actually happens when a senior finance professional leaves and the role goes unfilled on the theory that AI covers it. AI can generate the reports that person used to produce. What it can't do is own the judgment that person applied: knowing when a number looks wrong, deciding whether an anomaly is a timing difference or a real problem, and taking responsibility for what gets signed off. Even in heavily automated finance workflows, exceptions still get "flagged for review" and someone has to own them. Remove the person who owned that review, and the automation has no qualified check.
It's easy to see the tasks AI absorbed and conclude the role is redundant. It's much harder to see the judgment, oversight, and accountability that left with the person, because those don't show up in a task list. In finance, that invisible layer is often the most important part of the job. A team that keeps not backfilling its judgment-heavy roles doesn't get leaner and more efficient. It accumulates unowned risk, one unfilled role at a time, until there's no one left who can actually govern what the AI is producing.
The report itself points toward the smarter move. Alongside the no-backfill trend, it documents companies reallocating toward "AI-native roles" and hiring for people who can build and deploy AI. The winning pattern isn't "don't replace anyone." It's "replace deliberately, with AI-fluent talent." Instead of leaving a judgment-heavy finance role empty, the companies pulling ahead are filling it with someone who combines that judgment with fluency in the tools, a person who can both do the higher-order work and govern the automation handling the rest. That's how you build a modern finance team: not by shrinking the team reflexively, but by making each hire count more.
Some roles can go unfilled because AI genuinely absorbed them. Others are load-bearing in ways that don't survive being left open, and treating those the same way is how efficiency turns into exposure. For finance leaders, the discipline is distinguishing between the two, and refusing to let "AI makes us more productive" become an automatic reason not to hire finance talent for the roles that still need a capable human.
When a judgment-heavy role does need filling, the goal is to fill it well, with AI-fluent talent that raises the team's capability rather than just restoring its headcount. That profile, deep financial competence plus genuine AI fluency, is hard to source quickly through conventional local hiring. This is why more finance leaders are widening the search. MAVI surfaces pre-vetted, globally sourced professionals so you can hire finance talent with exactly that combination. Through our proprietary matching intelligence, each role you fill is upgrades your team, instead of just filling closing a gap.
ICONIQ's data shows companies increasingly using AI as a reason not to backfill. For the right roles, that's smart discipline. For the judgment-heavy finance roles that hold the operating model together, it's a mistake that compounds. The finance leaders who get this right won't be the ones who reflexively leave every seat empty. They'll be the ones who replace deliberately, filling the roles that matter with AI-fluent talent capable of governing everything the automation now produces.
ICONIQ's 2026 State of Scaling report describes companies increasingly using attrition "as an opportunity to avoid, delay, or down-level backfills," with no-backfill policies "explicitly tied to AI-driven productivity improvements." When someone leaves, the role often isn't replaced, on the theory that AI has made the remaining team productive enough to absorb the work.
No. For repetitive, high-volume work that AI genuinely absorbs, a no-backfill decision can be a legitimate efficiency move. The report documents real productivity gains, with 70%+ of employees as daily AI users. The mistake is applying the reflex to judgment-heavy roles that the operating model still depends on.
Because AI can reproduce the tasks a person did but not the judgment they applied, knowing when a number is wrong, owning exceptions, taking responsibility for sign-off. The report shows even automated finance workflows still flag exceptions for human review. Removing the person who owned that review leaves the automation with no qualified check.
Not yet, according to ICONIQ. The report finds functional headcount mix has stayed largely unchanged and concludes that "while AI is improving productivity, it has not yet meaningfully reshaped operating models." AI is making existing people more productive rather than replacing the underlying structure of who does what.
Replacing deliberately rather than reflexively. The report shows winning companies reallocating toward AI-native roles and hiring people who can build and deploy AI. For judgment-heavy finance roles, that means filling them with AI-fluent talent, letting leaders hire finance talent that upgrades the team rather than leaving a capability hole.
