Brief #1: Before You Launch Another AI Pilot, Do This First
Welcome to the Chief of AI Brief. Every Tuesday, we share what’s working, what’s failing, and what’s coming next in AI for business. Let’s start with the biggest blind spot most companies miss.
If someone shared this with you or you came from LinkedIn, subscribe to Chief of AI Collective to get your Briefs directly in the inbox.
McKinsey recently quantified what we've seen everywhere: 78% of companies are using generative AI, but 80% report zero material impact on the bottom line. Most are drowning in "horizontal" pilots like copilots, chatbots, and plug-ins, while truly transformative applications ("vertical," function-specific use cases) remain stuck in endless testing.
McKinsey calls it the gen AI paradox and argues companies must reset their AI transformation approaches from scattered initiatives to strategic programs.
Great advice, but how do you actually do it?
💡 Step One: Appoint a Clear Owner
The first move, and the one most companies skip, is assigning someone to own AI implementation. Not to build or use models, but to oversee adoption: choosing tools, tracking performance, and evaluating ROI.
What we’re seeing across companies is CEOs and managers asking teams to experiment with AI tools and report back. But without a clear owner, those efforts often stall.
That’s why over 37% of large U.S. companies, from Walmart to Johnson & Johnson to J.P. Morgan, already set up what’s called an AI Center of Excellence. You don't have to be an enterprise to start one too. It can start as a small team or one dedicated person (your Chief of AI) who is responsible for moving implementation forward.
Why does dedicated leadership matter? You don't ask employees to evaluate their own job performance (HR does that), and you don't let departments choose their own software without IT oversight. AI implementation needs the same cross-functional supervision, especially because the decisions you make now will compound over time.
Read more about these cases:
Walmart Case Study: Best Practices for Setting Up an AI Center of Excellence (CoE) in Retail
How AI Centers of Excellence are transforming global operations at Johnson & Johnson
AI Center of Excellence (AI CoE): Meaning & Setup in 2025 (Use Cases: J. P. Morgan, Siemens, IBM)
🛠️ What This Team Actually Does
First 30 days: Audit AI use across every department. Build a simple inventory: tools in use, costs, results, and whether ROI is being measured.
Categorize everything into three buckets:
Expand: Tools showing clear business results that could scale
Fix: Promising tools facing specific obstacles
Stop: Tools that haven't delivered results after six months
Next 60 days: Focus on quick wins. Take the "expand" category and figure out how to scale successful pilots to other departments. Take the "fix" category and identify specific barriers—usually data quality, system integration, or user adoption.
👤 Who Should Own This Role
Background that works: Operations, strategy, or business process improvement. Someone who understands both how work gets done across departments and the AI technologies that can transform it.
Background that might not work: Pure IT or data science. Technical skills matter, but this role is more about business transformation, not model building.
Key qualities:
Can talk to department heads and understand their real pain points
Comfortable with both technology and business operations
Has credibility across the organization to ask hard questions about ROI
⚠️ Common Mistakes to Avoid
Making it just an IT project. IT can support AI implementation, but they shouldn't own strategy decisions about which business processes to target.
Choosing someone too junior. This person needs to push back on department heads who want to keep ineffective pilots running. That requires organizational credibility.
Choosing someone too senior. C-suite executives often lack the hands-on tech fluency needed to evaluate emerging AI tools and implementation challenges.
Starting with tools. The first question isn't "What AI tools should we use?" It's "What problem are we solving?"
After all, AI transformation isn’t about adding tools. It’s about changing how work gets done. Assigning clear ownership early is the fastest way to turn scattered pilots into real business impact.
💬 Have you created this role or are trying to? We'd love to hear from you.
Thanks for reading,
Chief Of AI Collective
Interesting perspective on appointing one person but what about embedding AI use and understanding with leadership themselves and what happens to their role as critical change agents? How does this work?