Find, Fill, Fix: How AI Pressure Is Changing The Ask

Leaders are being asked to do more with leaner teams, and the AI question lands on everyone's desk at once. Across our executive search, interim leadership, and team performance work, we're seeing that pressure show up in three places: who leaders ask us to find, which gaps they need us to fill, and what they want us to fix on their teams.

Charter's new AI Pulse, a survey of 5,002 US desk workers, managers, and executives who use AI, found that 42.7% say AI makes their work and life at least somewhat better. Only 10.5% say it makes both much better, which the researchers define as thriving. More output was never the right measure. Two things matter: whether the right work is getting done, and whether the experience of that work is sustainable and, frankly, even enjoyable (gasp!) for the team doing it.

Here’s the one idea that ties our three practice areas together: AI can take on more of the work, but a person still owns the outcome, so the people carrying it need to be the right people, with coverage and capacity.

Find: the Head of People role now asks for a rare mix of skills

The Head of People role has evolved into a rare blend of skills. People and culture are still the foundation, but the role now stretches to include AI and analytics fluency, and few candidates have all of it.

Here's what we think is behind the shift. McKinsey's latest survey, reported by Charter, found 80% of respondents saying AI improved their individual productivity, while only 37% report enterprise-level financial impact. BCG's Julie Bedard makes the same point: the real gains come from reinventing team and organization-wide workflows, not from personal productivity. That takes a leader who can read the data and guide the change.

Trust turns out to be a performance lever. In MIT Sloan Management Review, Charter's Brian Elliott and workforce adviser Sophie Wade cite i4cp research finding that organizations whose people trust their leaders are 11 times more likely to be high performers. The technical skills get the headlines, but the human ones carry the weight, which is why finding someone with all of it is so hard.

We've also been filling Chief of Staff roles. Leaders recognize they often need a body double, someone who protects time for strategic work on top of the month-to-month deliverables. For example, when a large health system’s CEO asked the CPO to turn her focus outward, we placed a Chief of Staff with deep HR experience to take on the internal tactical work. That freed the CPO to lead and build trust across the organization, their customer community, and her team.

Fill: gaps keep opening, and leaner teams can't absorb them

With leaner teams, any kind of gap in coverage means backfilling becomes essential. (Babies continue to be born!) A year ago, teams could absorb an absence. Now, with a key person out, the work stalls.

We're also exploring something newer. Sometimes a full-time hire isn't needed, and our interim experts cover just the portion of work required (e.g., sales compensation updates). We're curious what happens when we pair an interim sales comp expert with our AI consultant to build an agent that keeps the work running after the expert rolls off, with a named owner on the client side. Each year, a client re-engages a compensation consultant to run the full end-to-end cycle from scratch. Imagine an agent that refreshes the data and flags gaps, so the work keeps running between engagements, allowing the consultant to play a more strategic role.

This could work across many interim skill sets. 

The handoff is where this succeeds or fails. Anthropic's Kristen Swanson, quoted by Elliott at Charter, says the teams that work best give their agents a clear role, written context, and goals set by a human owner. An interim engagement is a natural moment to write all of that down.

A good handoff comes down to three questions we'll return to below: who owns the agent, what is it allowed to do, and where is the work documented? What work do you have that isn't enough for a full-time employee, but where a hybrid like this could meet the need?

One caution: the AI Pulse found that daily use matters more than advanced use like building agents. Our goal here is continuity, and an agent only helps if someone owns it and uses it.

Fix: manager capacity is the constraint most teams feel first

Almost no one is only managing anymore. Managers deliver their own work and also guide their teams on using AI well, and those two jobs compete for the same hours.

That matters because managers shape how AI lands. Charter's AI Pulse found that 11 of the top 14 factors tied to benefiting from AI are manager behaviors. It also found a demonstration gap: managers are about twice as likely to say they do the recommended things as employees are to recall seeing them. We suspect part of that gap is capacity, since manager-doers have less room to model new habits. That's our hypothesis, not the researchers' finding.

Two more findings add to the picture:

• Training: Fewer than 40% of AI users receive the hands-on training they value most, and 21.4% say their organization offers none. In a BCG case study cited in MIT SMR, AI use rose 89% when managers tailored training to each person's adoption mindset.
• Hiding: 36% of AI users have hidden their use from a manager, colleague, or client, mostly to avoid looking incompetent. Managers are 1.6 times more likely to do it than workers, and it's the top factor tied to burnout.

The takeaway: managers shape whether AI helps, and they need the capacity to do it.

Engineering is speeding up. Communication is the bottleneck.

At The Lobby, a conference bringing together VCs and founders, we heard a trend worth sharing: AI is accelerating engineering, but marketing and communications can't keep pace in telling customers what changed. Customers, meanwhile, can only absorb so much improvement at a time.

Shipping faster only creates value if customers understand and adopt what shipped, so the pace of release may need to match customers' capacity to absorb change, not engineering's capacity to deliver it.

Whoever owns a change also owns helping customers understand it.

Where it's clearest: who owns the agent

Clients are debating whether agents belong on the org chart. A survey of more than 1,200 managers by BCG found that about 31% of organizations already frame AI as a teammate or employee, and 23% list agents on the org chart. What's clear, either way, is that accountability always sits with the person who manages the agent.

Brian Elliott's recent Charter column separates two decisions hiding inside the idea of putting agents on the org chart. Giving an agent a clear owner and a defined job is a good idea. Giving it a human name appears to erode accountability. In the same research, managers who reviewed identical error-filled documents caught fewer errors when told the work came from an AI employee rather than an AI tool, and they were more likely to pass it along for further review. The effect was strongest at companies already treating agents like staff. If you do name an agent, Elliott suggests keeping the name plain and descriptive.

Here's a simple test to bring to your next leadership meeting:

1. Who owns it? One named person who checks the agent's work and doesn't defer to it, not a team or a tool.
2. What is it allowed to do? Scope, limits, and what it must escalate.
3. Where is that written down? An agent can't act on context nobody documented.

What we don't know yet (well, there is too much of that to fit on a blog...)

We're still learning how to size a manager's workload once agents are in the mix, because counting direct reports is so 2025. We're also watching whether the manager capacity squeeze eases as tools mature or simply shifts. We'd love to compare notes with you.

Where this leaves us

Across all three practice areas, the opportunity is the same: make sure the right people own the work, with the coverage and capacity to do it well and keep it sustainable.

A few ways to take the next step, at whatever level fits:

• Start small: Run the three-question ownership test on one agent or workflow this week.
• Talk it through: Book a 30-minute conversation about your Find, Fill, and Fix priorities.
• Join the conversation: Tell us where in your organization the work is most at risk of having no clear owner. We're sharing more on LinkedIn and would love to hear your thoughts.

We're excited to keep exploring this with you.