Reading · Ask a Chief of Staff

If you're reading this as a forward or from LinkedIn, subscribe to Ask a Chief of Staff to get these insights directly in your inbox. Join 16,000+ Chiefs of Staff who trust us for actionable insights.

Every company I speak to right now seems to be in the middle of a Claude rollout. If this is you, I know some guys you should talk to.

I’ve known the founders from Ambient for a few years, and I’ve been consistently impressed with their depth and thoughtfulness around how companies can go AI Native.

They’ve just launched a new Claude-focused consulting business called, Legible.co. Their belief is that for companies to truly lean in to AI, they need to become “legible” to AI, which often requires changes in policy, operating model, and a general upleveling of skills - all things the Chief of Staff is often in the middle of.

The core offerings of Legible include:

Claude trainings and enablement
AI strategy
Workflow automation and custom agent builds

🔥 Exclusive offer for Ask a Chief of Staff Readers: The Legible founders have kindly offered a free 3️⃣ 0️⃣ minute Claude training session (individual or group). Just include AACoS26 when you fill out the lead form.

Talk to Legible

*If you’re interested in sponsoring a future issue of Ask a Chief of Staff, hit reply!

This issue is guest authored by Sydney Gorski, Chief of Staff at Collide, a VC-backed startup building AI for energy operators. After nine years planning events and meetings, she found her calling in the magic middle of the Chief of Staff role: where strategy and ops meet people and culture. Turns out running complex events is excellent training for running a company.

Everyone’s racing to become AI-native. Almost no one’s asking what we lose along the way.

That’s our job. Chiefs of Staff are the keepers of company culture, and in the AI race, that job just got harder.

There’s a line I love about this role:

Chiefs of Staff have influence over everyone and authority over no one.

Nowhere does that play out more than on the culture side. We sit at the intersection of every department. We hear what’s said in the leadership meeting and what’s said in the hallway after. We’re usually the first to notice when someone’s mood is off, when someone has one foot out the door, or when someone is stuck and needs encouragement.

Most of us are wired to read people well. That’s not incidental. It’s why this role works.

This matters more than ever right now because everything is moving fast. Companies racing to become AI-native are rolling out new tools and workflows on weekly cadences.

AI is intuitive enough that people assume it’s easy to roll out, and they skip the questions that should come before any major change, like:

❓ How will this affect our culture?

❓ What events, what processes, what rituals make the company what it is?

❓ What are we taking away?

That’s where the Chief of Staff comes in. People look to us to set the temperature for the company.

We’re often the bridge between a leadership team that can feel distant and the rest of the organization. In the age of AI, the bridge job has new weight. We’re the ones who translate between speed and sustainability, between adoption and adjustment, between the excitement of new tools and the discomfort that comes with them.

When we talk about protecting company culture in the face of AI, there are 3️⃣ risks worth watching closely.

The biggest risk AI poses to culture is silos.

You can see it forming on almost any team that’s gone AI-forward. Engineers are building tools their teammates don’t know about. Ops folks are spinning up email agents in parallel. Three different people are building three different versions of the same workflow agent. AI is great, but it needs direction and purpose.

Letting everyone run wild gets you people going rogue, building things that aren’t actually valuable just for the sake of building them, and 20 versions of what should have been one tool.

The leadership team is just as exposed. Strategic planning is a perfect example, and AI can absolutely make the process more efficient.

But what gets lost is the camaraderie of being a leadership team together. What’s the point of being on a leadership team if Claude is the one calling all the shots? 🤖

AI is only as good as the information you feed it, and there’s a lot it can’t see. It can’t do the accountability chart work, the right person right seat conversation. It can’t capture the context a teammate picked up at a drink with a client last week, or the read on someone who’s been quieter than usual in standup.

Strategic planning should never live solely in Slack or Microsoft Teams. The work that builds a leadership team can’t happen async.

When team members can ask AI for advice, the work gets done.

What gets lost is everything that used to happen around the work: mentorship, peer learning, the back-and-forth that turns one person’s question into the team’s shared knowledge.

AI doesn’t have context of how a manager solved this same issue two years ago, or what the team learned the hard way last quarter.

There’s a parallel here to “building in public,” the social media strategy where founders share their work openly to grow an audience. The same idea works internally as an operational strategy. When teammates build with AI in private, in their own Claude window or their own DM with the model, the work stays siloed and the learning stays trapped. When they build in public, in a shared channel where their thinking is visible to the rest of the team, the work compounds and others learn from it.

The problem might be solved when someone goes to AI alone. But did they learn anything? Did anyone else? 🤔

This one is the hardest. In the US, we say, “I am a Chief of Staff.” In Europe, they say, “I work as a Chief of Staff.” Our jobs are our identities.

So when AI starts automating big chunks of someone’s role, it’s not just a productivity question. It’s: “Who am I without these things I’ve been doing every day for the last couple years?” 😬

This hits hardest for the operations and admin folks, the people whose jobs are made up of small daily tasks. Of course we want to automate those tasks. But the person doing them can’t help but ask: if my entire job is automated, why do I still have a job? AI stops feeling like a thought partner and starts feeling like a replacement.

That fear is real, and it is not being talked about enough.

This is where the role earns its keep. Here’s what Chiefs of Staff can do to protect company culture in the face of AI, starting now.

Most teams aren’t talking honestly about how AI can make people feel irrelevant.

It’s all so exciting, getting more done, having more ideas. But the discomfort is real, and silence makes it worse 🤐

As Chiefs of Staff, we can be the ones who name it. Help people see the human value in how they show up to the workplace, not just the work product they ship. Make sure they know automating the repetitive parts of their job is a good thing because it frees them up for the things AI can’t do. Then make sure those things actually exist for them to do.

This is the discipline that needs to become muscle memory for every Chief of Staff. Before any new AI tool or workflow gets rolled out, the team should be answering three questions:

1️⃣ If we automate this, what conversation goes away?

2️⃣ If this is replacing a meeting or workflow, who used to be in the room, and what did they bring beyond the work?

3️⃣ What signal was this process giving us that we’ll lose if we automate it and stop questioning it?

Not everything is worth automating. Manual processes still have a time and a place. When we think about onboarding a new employee, new hires can absolutely work through mundane company policies in a self-guided AI tool on their own time, even before their start date. That kind of low-stakes content is exactly what AI is good for. But someone still needs to sit down with every new employee in person to go over what it actually means to work at your company.

A company is only as good as its values, and core values, beliefs, and mission are the foundation of company culture. Those can’t be writing on the wall or part of an AI onboarding flow.

Asking the three questions doesn’t mean you’ll get the answers you want. The CEO might still want to ship the AI tool tomorrow. The leadership team might roll their eyes at the culture concern. Two things help.

✅ First, translate culture into operational language.

“I’m worried about culture” doesn’t move a fast-moving CEO. “I think we’ll pay a rework tax because adoption will stall” does. Frame the cultural cost as an operational one. Speed people care about speed. Show them the friction they’ll create by skipping the human work, the time they’ll lose unwinding sloppy rollouts, and the meetings they’ll spend explaining tools that didn’t land.

✅ Second, always beta test.

When you can’t stop a rollout, slow it down. “Let’s run this with one team for two weeks before we go company-wide” almost always gets a yes because it sounds like one. The pilot data will tell the story you couldn’t tell with words. If the pilot goes great, you’ve helped the rollout. If it goes badly, the team sees the friction themselves and you didn’t have to be the one to predict it.

✅ And keep coming back to the 3️⃣ questions.

Drop them into the conversation as a shared team exercise, not your personal objection. The conclusions land harder when the leadership team gets there together.

There are touch points AI can’t replace, and we have to actively protect them.

AI can’t make someone feel special on their birthday. AI isn’t going to remember that someone is thinking about proposing or that someone just bought a new house. Those are conversations that happen during drinks after work, on a walk over a lunch break, or in the car driving to a client meeting. Those moments are what make a workplace feel like one.

A few rituals worth protecting and building:

💫 The segue. Start every meeting with everyone sharing a piece of personal good news and a piece of business good news. It’s an EOS practice. It humanizes everyone and reminds the team that we’re not just worker bees prompting AI all day.

🙋 Monthly show-and-tell. Make space for anyone across engineering, product, marketing, ops, or content to share something they’re working on. Depending on the size of your company, this can be company-wide or team-wide. New product, new content, new metrics, a new app. It humanizes the AI everyone is building. It’s one of the easiest ways to break silos and one of the most reliable culture builders I’ve seen.

🏗️ A dedicated AI build channel. Create a Slack or Teams channel where team members can share what they’re building with AI, test ideas with each other, and get feedback before they ship. This is what building in public looks like internally, and it catches duplicative work before three people have all built the same thing.

💼 Monthly AI office hours. Host them with whoever on your team is leading AI enablement. Open the door for anyone to bring questions, blockers, or half-built ideas. This is especially valuable for the less-technical folks on your team who need a low-stakes way to get up to speed.

🎉 Sharing wins in Slack. Small but it adds up.

In-person interaction matters more than ever 🧑‍🤝‍🧑 Not because we want to spend time together, but because we need to make sure we don’t lose the human elements of what makes a company special. This is harder for remote teams, where the casual moments don’t happen on their own. You have to build them in.

It is our job to fight for our culture and protect it at all costs.

It’s easy for leadership teams in fast-moving companies to put blinders on and go all in on AI. You can see what happens when AI tools get rolled out company-wide without real onboarding or thought: adoption stalls, the team gets confused, and the promise of efficiency ends up creating more friction instead of less.

This is where the bridge work matters most. We sit between what’s realistic and what’s not, between leadership and the rest of the team, between the technical folks running ahead with AI and the non-technical folks worried about being left behind.

Our job is to translate between them, to flag what’s at risk, and to fight for the rituals, the conversations, and the moments that make the work mean something.

In the age of AI, that’s not a part of the job. It is the job.

Thanks for reading! This post is public so feel free to share it.

Ready to accelerate your growth in 2️⃣ 0️⃣ 2️⃣ 6️⃣ ? Paid Substack members get exclusive access to a growing library of 2️⃣ 0️⃣ ➕ workshops from industry experts—covering everything from strategic planning, to navigating your next career move, to keeping cool under pressure 🧘‍♂️

New workshops added monthly! Recent additions include:

$1️⃣ 0️⃣/month or $1️⃣ 0️⃣ 0️⃣/year. Cancel anytime.

Join 300+ Members

Save Your Seat Now! ☝

June 11th: AI for the Chief of Staff: Using Claude to Automate Repetitive Tasks

June 11th: Peer Resume Review

June 16th: Where To Start When Your Boss Says: “Figure out AI”

June 23rd: Chief of AI Strategy: The Playbook for Advising on AI Adoption

June 30th: From Slide Decks to Shipping: What Consulting Didn’t Teach Me

June 10th: Ask a Chief of Staff x Offsite: Atlanta Community Dinner

June 11th: How a Chief of Staff Operates Across Stage & Scale (Atlanta)

June 17th: NYC Co-Working Day at Industrious