.avif)
.avif)
Here's the thing nobody wants to say out loud: using AI every day doesn't make you an AI leader. It makes you a power user.
And in 2026, those are not the same thing.
What 250+ Chiefs of Staff actually told us
We ran the COSN AI Readiness Diagnostic and pulled a sample of 250+ senior leaders from the results to compile this report, drawing from Chiefs of Staff and senior operators across technology, financial services, logistics, media, and professional services. This isn't a consumer poll. Think of it the way Gartner or McKinsey frames their C-suite panels: a specialized executive sample, built for depth, not breadth.
The results were clarifying, and a little uncomfortable.
The surface numbers looked great:
- 86% use AI tools at the highest frequency levels, daily, across multiple tools
- 90% have used AI to draft internal communications in the past 30 days
- 70% have used AI for board decks or executive briefings
- Nearly everyone is in ChatGPT, Claude, or Perplexity every single day
By every surface-level measure, the Chief of Staff community is AI-forward.
But dig one level deeper:
- Only 7.3% scored as AI-Native, meaning operators who have built systems where AI runs continuously, surfaces insights they'd otherwise miss, and manages parts of the workflow without them
- Only 12.5% connect AI to other tools via APIs, automation platforms, or custom integrations to create multi-step workflows
- Only 5.2% have AI agents actively managing parts of their project workflow
- 15.3% said strategic planning is still a fully manual process with no AI involvement at all
The other 92.7% are using AI. They haven't become Chief of AI. That gap matters more than almost anything else happening in this role right now.
The framing we've been given is wrong
The dominant narrative around AI and the CoS role goes something like this: learn to prompt well, adopt a few tools, and you'll stay ahead. It positions AI fluency as a skill to acquire, like Excel proficiency or slide mastery. Get good at it and move on.
That framing is too small. And for Chiefs of Staff specifically, it misses both the opportunity and the risk.
The Chief of Staff sits at the intersection of strategy and execution. You see everything. You talk to everyone. You're the person whose job is to make the whole system work better. That position is uniquely powerful for AI leverage, and uniquely vulnerable if someone else claims it first.
Many Chiefs of Staff are in a position to lead AI transformation at their organizations right now. And not enough are actually doing it.
Think about where you sit. You have more context about where the organization is inefficient, where decisions are slow, and where the right information isn't getting to the right people than almost anyone else in the company. That's an enormous amount of leverage for driving AI adoption from the inside. Our 2025 Chief of AI Report found that Chiefs of Staff are the second most common decision maker when organizations choose which specific AI tools to implement, trailing only the CTO and ahead of the CEO. The seat is already there.
But positioning yourself as your organization's AI leader isn't something that happens to you. Nobody hands out that mandate. The CoS who are becoming Chief of AI are the ones who picked up the agenda themselves, ran the audit, built the internal case, got their CEO aligned, before anyone asked them to. That's the opportunity most of the network isn't taking.
What "Chief of AI" actually means
It's three things at once, and they compound.
First, it's a mindset shift. The operators who scored AI-Native aren't just using better tools. They think about AI differently, not as a productivity layer on top of their existing workflow, but as a structural change to how work gets done. They ask: where is human judgment genuinely required, and where am I doing something AI could do better and faster? That question, asked honestly, is uncomfortable. It's also the right one.
Second, it's a specific set of capabilities. 54% of respondents write detailed prompts and iterate on outputs, and that's real skill. But there's a big difference between prompting well and building systems. Connecting AI to other tools via APIs, creating multi-step automations, deploying agents to manage recurring workflows, that's where the ceiling is, and most people haven't gotten close to it.
Third, it's a strategic portfolio. Becoming Chief of AI means taking ownership of your organization's AI direction the same way you'd own any other strategic initiative. Running a capabilities audit. Evaluating and sequencing tool adoption. Building internal fluency programs. Managing the risks around hallucination, data exposure, and over-reliance. And briefing your principal with a point of view, not just a summary of what's out there.
Strategic planning is exactly where the Chief of AI role has the most to offer, and the fact that 15.3% of respondents still haven't touched it with AI at all is the single biggest gap in the data.
The tools Chiefs of Staff are actually using
The AI tool stack among senior operators has sorted itself into three distinct tiers. The distance between them is wider than most people realize.
Tier 1: The core stack (commoditized, universal)
These are the tools every CoS is already using. Awareness is near-100%. The question isn't evaluation, it's how deeply they're being used.
- ChatGPT (OpenAI): still the most widely adopted general-purpose AI across the community
- Claude (Anthropic): rapidly growing, especially for longer-form analysis and writing
- Perplexity: the go-to for research and synthesis, consistently mentioned alongside ChatGPT and Claude
- Microsoft Copilot: high awareness, lower adoption in smaller orgs, and still underutilized relative to how broadly it exists in M365 environments
- Google Gemini: present for Google Workspace users, but rarely the tool of choice
Tier 2: The productivity layer (where differentiation is happening)
This is where AI-Native operators are pulling ahead. These tools are known by many but actively evaluated and used by significantly fewer.
- Granola: AI-powered meeting notes that are growing fast in this community as a step up from Otter.ai and Fireflies
- Otter.ai / Fireflies: widely known, but most CoS haven't moved past basic transcription to actually using the AI analysis features
- Gamma: AI presentation builder used by a subset of operators who have stopped fighting with PowerPoint
- Notion AI: there's significant Notion adoption in this community, but the AI features remain underutilized relative to the base platform
- Superhuman: known but less actively evaluated than its level of awareness would suggest
Tier 3: The integration and agent layer (the actual frontier)
Only 12.5% of the network has reached this tier in any meaningful way. These are the tools and capabilities that separate AI users from AI-Native operators.
- Zapier / Make: automation platforms that connect AI outputs to other systems, widely heard of, rarely built into daily workflow
- Custom GPTs / AI agents: 5.2% of the network has actually deployed these into their workflow; for most, it's still theoretical
- n8n, Relay.app: nearly invisible in this community despite being built for exactly the kinds of workflows CoS manage
What this means for AI companies
The Chief of Staff is not just a user of AI tools. In most organizations, they are one of the key people deciding which tools get evaluated, piloted, and purchased. Our 2025 Chief of AI Report confirmed this: the CEO-CoS-CTO triad is how companies are making AI decisions, with the CoS serving as the executor, analyzing the problem and projected impact, and project managing the buying and implementation process.
Tier 1 brands have the awareness. Tier 2 is where the market is being won right now. Tier 3 represents the biggest untapped opportunity: tools purpose-built for how CoS actually work, but that haven't yet cracked the discovery problem with this audience.
For brands in Tier 2 and Tier 3: the COSN audience isn't evaluating you yet, not because your product isn't relevant, but because you haven't shown up where this community learns. That's a marketing problem, not a product problem.
The gap isn't tools. It's intentionality.
Every operator in the network has access to the same Tier 1 stack. ChatGPT, Claude, Perplexity, the core is basically democratized. What separates the 7.3% isn't access. It's intentionality.
AI-Native operators have made a decision, explicit or implicit, that they're going to figure this out ahead of the people around them. They're not waiting for their company to develop an AI strategy. They're developing it. They're not waiting to be asked to lead on AI. They're already leading.
That gap compounds fast. The operators who build AI-native workflows in 2026 will have a structural advantage in 2027 that's genuinely hard to close. Not because the tools will change, though they will, but because the judgment, the pattern recognition, and the institutional trust that comes from doing this work early doesn't transfer. It accrues.
What comes next
If you scored as a Practitioner or Experimenter in the diagnostic, that's not a criticism. It's a starting point. The majority of the network is in that zone. The question is what you do next.
For Chiefs of Staff who want to close the gap and actually become Chief of AI rather than just use AI well, the place to start is doing the work alongside operators who are already there. That's what the COSN AI Fellowship is built for.
The Fellowship brings together Chiefs of Staff who are serious about building AI into the core of how they operate: their workflow, their strategic toolkit, and their leadership approach. It's structured, peer-driven, and built specifically for the CoS context, not generic AI training for enterprise employees, but applied AI for operators who already understand the terrain.
If you took the diagnostic, you already know where you stand. The next question is what you're going to do about it.
Also coming up: AI & The Law
Becoming Chief of AI isn't just about building workflows and deploying tools. It's about being the most informed person in the room when your organization is making AI decisions, including the ones that carry legal and regulatory risk.
On May 19, 21, and 26, COSN is running a 3-session live course with experts from Clio, Orrick, and Tech Contracts Academy covering exactly that territory: AI risk and adoption strategy, the current state of AI governance (EU AI Act, CCPA, and what's already being enforced), and how to actually read and negotiate AI contracts before your organization signs something it shouldn't.
If you're the person your CEO is going to turn to when these questions come up, this is worth your time.
Register for AI & The Law (enrollment closes May 15)
The AI Readiness Diagnostic is available to all COSN members and prospective members. Take it at chiefofstaff.network/ai-readiness.
