What AI-First Leaders Do Differently

This Week

If we worked on your AI rollout in real life, I’d probably annoy you just a little.

Not with big vision talk or flashy demos.

Just by constantly asking: “Who’s doing this work?”

Because every time I sit down with a team that’s struggling with AI adoption, it’s the same pattern. The strategy is solid and the tools are in place, but no one really owns it, and no one has time. It’s just another bullet on someone’s Q4 plan, wedged between “launch campaign” and “respond to that Slack from HR.”

That’s when pilots stall, tools get ignored, and the whole thing starts to look like a mistake.

The teams that win?

They carve out time, assign people to the project, and build systems that fit their work flows. 

That’s what this week’s newsletter is about: what it really takes to implement AI inside a business successfully, sustainably, and without driving your team insane.

Here’s what we’ve got:

• The Companies Getting AI Right All Have One Thing in Common

• Why Friction Leads to Better AI Roll Outs

• Unlocking the 5 Barriers to AI Adoption

• How the Top AI Leaders Are Winning on AI ROI

• Using Influence Networks for Successful AI

• Why Most AI Pilots Fail: How to Make Yours Work

Oh, and I’m off to Italy for the next two weeks. This time I’m actually going to relax. No AI-generated selfies needed.

See you in late October.

Now go read it.

Teams Getting AI Right Have One Thing in Common

It’s not better models or bigger budgets. And it’s not that they hired a chief AI architect.

The companies getting results from AI have something much harder to replicate: leaders who know how to translate technical possibility into business value.

These are the people who have built organizational muscle. They’ve stopped treating AI as an IT project and started treating it as a core part of how they and their teams operate.

They are what researchers Broek, Hellauer and Wang call AI Shapers. 

“The CEO must model curiosity and signal that AI is central to strategy; the CFO must reimagine financial processes; the CHRO must rethink how talent decisions are made. Shapers must be embedded across teams, functions, and levels.”

Broek et al. found that leaders who drive measurable returns from AI share five key capabilities.

They are:

Strategic agility
Leaders who excel in this area prioritize options over rigid plans and proactively scan for disruption before it derails progress. They focus on business value rather than novelty, avoid sunk-cost traps, and balance timing with speed.

Question to ask yourself: Do we have clear criteria for when to pivot our strategy, or are we locked into plans that may no longer serve us?

Human centricity
Adoption depends on trust. People need to know why AI is being introduced, how it will affect their work, and who to talk to about their concerns. Leaders strong in this area build psychological safety through empathy and feedback loops.
Questions to ask yourself: How can AI make humans better? Are we designing with employees in mind or just hoping they’ll figure it out?

Applied curiosity
The best leaders learn the work flows, processes and tools themselves. They test use cases, explore limits, and model continuous learning for their teams.

Questions to ask yourself: Are we solving a real problem? Am I actively engaging with AI or just delegating the learning to someone else?

Performance drive
Progress is measured by results, not effort. Leaders define clear goals, track outcomes, and stay focused on business impact.

Question to ask yourself: Are we measuring results that matter or just checking boxes?

Ethical stewardship
Responsible AI needs transparency, safety checks, and clear guardrails from the start. Leaders who take this seriously design governance that manages risks while accelerating progress.

Question to ask yourself: Are we building for accountability or hoping risk won’t catch up with us?

As the researchers say, some people naturally exhibit stronger capabilities than others, but the good news is that all of the capabilities can be developed. The key to doing that is to understand where you and your team stand today, what the biggest gaps are, and how to build systematically from there.

Want to read the HBR article? Head over here.

Why Friction Leads to Better AI Roll Outs

A recent MIT study found that 95% of GenAI pilots fail, mostly because companies try to sidestep friction rather than lean into it.

To be fair, the study’s methodology has raised a few eyebrows, but the core takeaway holds up: the 5% who succeed follow a remarkably consistent playbook. They don’t get distracted by splashy demos or throw generic copilots at every problem. Instead, they build AI systems that learn, admit when they’re wrong, and fit well into the work people are already doing.

According to Jason Snyder, tech writer for Forbes, here’s what winning orgs do differently:

• They build learning loops. Context retention and feedback are baked in, so GenAI gets smarter over time.

• They don’t pretend AI is infallible. Their systems admit uncertainty and learn from human corrections.

• They partner with specialists. Teams using external vendors with domain fluency and shared accountability are twice as likely to succeed compared to those building everything in-house.

• They focus on back-office tasks. That’s where the friction is highest and so is the return.

The companies that figure out how to work with friction (not get stuck in it) are the ones that’ll see real ROI. But they’ll also build something more durable: resilience, momentum, and long-term advantage.

Unlocking the 5 Barriers to AI Adoption

If AI still feels out of reach it’s probably not the tech that’s the issue. These four blockers come up again and again, whether you're running a team or flying solo.

Barrier 1: Low data quality
If your systems are full of duplicates, disconnected tools, or outdated customer data, AI won’t save you — it’ll just surface the mess faster.

Fix it: Start by cleaning what you’ve got. Standardize your inputs. Connect your tools. Even basic improvements to your data can unlock more accurate and useful AI outputs.

Barrier 2: Lack of knowledge
Many people still aren’t sure what AI actually does or how to use it in their role. That lack of clarity slows everything down.

Fix it: If you're solo, pick one use case (email, research, recaps, whatever) and start there. If you lead a team, invest in shared learning. Host a lunch and learn. Bring in a guest speaker (like me!). Do a show-and-tell where team members demo the AI tools they like. Small habits like these create a culture of curiosity and build momentum.

Barrier 3: Fear of change
No one wants to feel replaceable. And when tools start automating things people used to own, the reaction is often resistance, not excitement.

Fix it: Make it clear AI is here to support the work, not take credit for it. Give people space to explore, ask questions, and opt into the process early. Whether it’s streamlining research or writing better briefs, focus on the time saved, not the tasks replaced.

Barrier 4: No AI policy
According to a recent Section survey of 10,000 knowledge workers, 69% of employees are unsure about their organization’s AI policy and only 23% of employees report their company has a formal AI strategy. Without clear guidelines, teams are either unsure what’s allowed or taking unnecessary risks. One person might be uploading sensitive data into public tools, while another avoids AI altogether out of caution.

Fix it: Put a basic policy in place. Define which tools are approved, what data should stay private, and how AI should (or shouldn’t) be used in day-to-day work. If you’re solo, create your own personal checklist so you’re not winging it every time. 

Barrier 5: Limited access to AI tools
Over 80% of companies say they’ve piloted or deployed enterprise LLMs, but only 29% of employees in AI-forward organizations have access to them. Most teams are either given watered-down tools or cut off entirely.

Fix it: Make access the default, not the exception. Equip knowledge workers with enterprise LLMs, enable advanced features (like custom GPTs), and offer enough coaching to get started. Fast-track approvals for tools that drive output.

How Top AI Leaders Are Winning at AI ROI

AI pilots are easy. ROI is hard.

That’s exactly what this year’s Section AI ROI Conference set out to solve.

The top minds in AI shared how they’re turning experiments into real results and why 92% of early adopters say their investments are paying off. Some are seeing 4x returns in under 13 months, not from flashy demos, but from focused, well-integrated systems that solve business problems.

Across seven hours of content, data scientists, venture capitalists and AI experts, unpacked the strategies that are working from measuring AI to scaling across the enterprise. It’s a jam packed day!

If you're past the hype and ready to make AI useful, this is the playbook you want.

Watch the full conference on demand.

Using Influence Networks for Successful AI

When Alicia Abella, PhD, joined Novo Nordisk, she inherited zero AI infrastructure. 

She also discovered something common to AI leaders: the biggest barrier wasn’t technology, but psychology. Employees had the tools but were paralyzed by fear of regulatory violations.

Her breakthrough insight? 

Adoption happens through influence networks, not mandates. 

She shared her four steps for using virality and influence to drive AI adoption and ROI at the latest Section event:

1. Map the fear landscape first. Spend a few months understanding psychological barriers, not just technical ones. Fear stems from uncertainty, not capability gaps.

2. Build influence networks, not user bases. Create ambassador programs that turn AI enthusiasts into internal evangelists who can address peer concerns authentically.

3. Design governance for paranoia. In regulated industries, your intake process must satisfy legal/compliance teams first, business value second.

4. Market internally like a consumer product. Create posters, videos, town hall presentations. “This is no different than if I was launching the iPhone.”

Alicia’s success came from recognizing that AI adoption requires the mindset of a consumer marketer, rather than an enterprise IT rollout.

Why Most AI Pilots Fail And How to Make Yours Work

Most internal AI pilots fail because teams try to recreate what already exists… and do it worse.

If you’re running a pilot, don’t waste cycles rebuilding general-purpose tools. Focus your effort where off-the-shelf solutions fall short: domain-specific problems that depend on your unique data, workflows, and expertise. That’s where internal teams can win and where AI can create strategic advantage.

The most successful pilots follow a repeatable playbook:

1. Co-create with domain experts
Your SMEs need to shape your AI pilot from the start. Bring them into the build process and let them define what “good” looks like. Their input drives better models and builds buy-in.

2. Build where people already work
Pilots fail when they ask people to change how they work. They succeed when the AI is available inside the tools they already use, like Salesforce, Jira, SharePoint and Slack. If it’s in a separate tab, everyone will ignore it.

3. Offer multiple models
Let people choose the right model for the job. According to Eric Porres, Head of Global AI at Logitech, giving access to GPT-5, Claude, Gemini (etc.) inside a unified interface doesn’t overwhelm, it drives better outputs. In his experiments, users who started with short interactions were more likely to engage in longer, iterative sessions as they discovered different models’ strengths for specific tasks.

4. Tie it to strategy
AI ROI shows up as time saved, quality improved, or revenue unlocked, but unless your pilot is tied to a core strategic priority, it won’t get the leadership attention (or funding) to scale.

5. Plan for iteration, not perfection
Your first version won’t be right and it doesn’t have to be. What matters is how quickly it gets better. The best pilots are built to learn, so bake in feedback loops, track outcomes, and improve as you go. Treat your pilot like a system, not a one-off.

6. Define success upfront
No metrics = no momentum. Before you build, align on what success looks like and how you’ll measure it. Is it time saved? Fewer steps? Better output? If you can’t quantify the win, you won’t justify the rollout.

7. Staff it like a real product, not a side project
Your AI pilot can’t live off the side of someone’s desk. Successful projects have three key roles covered:

• A domain expert

• A technical owner

• A product/ops lead to drive rollout, adoption, and feedback

Ritu Chakrawarty, Gen AI Solutions Strategy at AbbVie says your first sign of a successful AI pilot will be behavioral: users demanding access rather than needing persuasion.

This "pull versus push" dynamic indicates whether you’re solving a real problem or just building an impressive demo that will eventually lose organizational support.

Want to Level Up Your AI Game?

If your team is ready for a hands-on AI strategy session, my custom-designed workshops are built to uncover the workflows that can save you hours every week.

Prefer to start small? My YouTube channel is packed with quick, practical “how-to” videos that show you exactly how I use AI tools for marketing, content, and automation.

Planning an event or conference? I deliver high-energy AI sessions that engage audiences and leave them with actionable strategies they’ll talk about long after the event. Book me for your event here.