Capable, but Not Confident

At the start of every cohort, I ask leaders to rate themselves on their AI capabilities. One question always sparks a lively discussion:

On a scale of one to ten, how confident do you feel about leveraging AI?

Most participants last Friday gave themselves somewhere between a one and a five. That led us to a useful thought experiment: what would it look like to be a nine or ten?

What usually emerges is the gap between capability and confidence.

Many leaders are already using AI more effectively than 99% of their peers, but they still feel like they’re somewhere in the middle of the pack. Why? Because the noise around AI, every new tool, every new model, creates the sense that you’re always behind.

Part of our job in the ​boot camp​ is “exposure therapy.”

By practicing real-world use cases and best practices, participants realize they can accomplish more than they think, and their confidence starts to align with their capabilities.

By the end of three weeks, most rate themselves around an eight, without a huge leap in technical skill, but with a completely different mindset.

The lesson is clear: AI mastery isn’t just about knowledge; it’s about confidence in applying it.

What AI Mastery Looks Like Today

This naturally raises the next question: what does “capable” actually mean in October 2025? What’s the baseline a business leader should expect of themselves?

Two useful data points:

1. OpenAI’s guidance on use cases.

When OpenAI advises companies on where to start, it highlights six areas where AI excels: content creation, research, coding, data analysis, ideation/strategy, and automation.

In ​client workshops​ we run, those six areas almost always surface.

Yet according to OpenAI’s recent “ChatGPT usage and adoption patterns at work” report, the majority of professionals ​still aren’t using AI consistently​ across even half of these:

So if you’re regularly using AI to write, brainstorm, analyze data, or speed up coding workflows, you’re already more than meeting expectations.

(article continues after this break)

(Article continues)

2. AI fluency expectations in leading companies.

One of the best case studies on AI adoption, Zapier drove implementation success through an AI Fluency Matrix with four levels, ranging from unacceptable to transformative.

In this model, what ‘capable’ looks like differs by role, for example:

• HR/People: save two hours per week by drafting interview guides and summarizing panels with ChatGPT
• Support: summarize tickets with ChatGPT and cite faster context shifts
• Marketing: draft first drafts of social posts and headlines with AI
• Product: draft PRDs, story maps, and synthesize user-interview notes with ChatGPT
• Engineering: use ChatGPT for simple coding tasks likeregex, unit-test stubs

Chief People Officer Brandon Sammut shared during a ​recent masterclass​ that leaders who hit just the “acceptable” baseline are already outperforming much of the market.

In other words: if you can demonstrate consistent use of AI in your daily work, align AI to your team’s workflows, and evaluate outputs critically, you’re right where you need to be.

Mastery as an AI Leader

But “up to speed” is not the same as mastery. Mastery means being ahead.

As I shared in my recent Beyond the AI Surface webinar, leaders who truly master AI go further. They display skills like:

• Advanced prompting: understanding how to SuperPrompt and direct frontier models like GPT-5 with precision.
• AI as a senior thought partner: not just asking AI for outputs, but using it as a sparring partner for strategy and decision-making.
• Workflow integration: starting with tasks, not tools, they build repeatable, potentially automated processes where AI isn’t just a tool, but a true “copilot.”

As shared before, Josh Bersin calls these AI-forward leaders “SuperWorkers.”

They don’t just use AI to be more efficient, but as a way to expand capabilities, allowing them to become more creative and strategic.

And while they might not be AI experts by training, they have enough knowledge to ask the right questions and to challenge their teams on AI opportunities.

Crucially, advanced leaders have overcome much of their initial skepticism and now trust their intuition and experience with AI enough to advocate for its broader use. We often see leaders at this level actively role-modeling the desired behaviors.

(Our next Lead with AI ​Executive Boot Camp​ cohort kicks off on November 14 for those who want to build both the basics and these advanced skills.)

This is crucial, as personal mastery is only half the story.

Driving Mastery in Teams

Being a true AI leader means taking the next step: scaling fluency across your team, department, and organization.

OpenAI’s recently released ​AAAAG Framework​ emphasizes the leader’s role in building not just capability, but culture.

According to the framework, leaders should:

• Align: Create strategic clarity through executive storytelling, organization-wide AI goal, and role-modeling (as we saw on full display in the Moderna case study.)
• Activate: Building skills and enabling experimentation through training programs, a champion network, hackathons and performance goals.
• Amplify: Scaling success across the organization through AI knowledge hubs, success storytelling, internal communities, and team-level recognition.
• Accelerate: Removing friction to move fast by giving access to tools and data, intake + prioritization processes, AI councils and rewarding innovation.
• Governance (they ran out of A’s): Oversee responsible AI use through clear guidelines, regular reviews, and evolving policies.

At the advanced stage, AI is embedded in multiple functions and workflows, even if it’s not yet ubiquitous. The organization has progressed from isolated use cases to a portfolio of AI applications across the business.

For example, in a bank, AI might help underwrite loans, streamline internal processes, and personalize marketing offers.

These initiatives are often backed by a more centralized AI infrastructure: the company may have established a Center of Excellence or dedicated AI team to support deployment. (See my need for an “​AI Implementation Sandwich​”)

There is also a stronger data foundation: enterprise data lakes or platforms that feed AI models to enable scaling.

As a result, over half of companies at intermediate maturity have begun to ​achieve quantifiable financial impact from AI​ (about 54% of European CEOs and 60% of US CEOs in a 2025 sample reported real AI-driven financial gains at this stage).

This is exactly where our ​custom client work goes​: mapping pathways for individuals, then scaling adoption to the organizational level.

Mastery, at this stage, is about your ability to make AI a shared competency across the business and turn it into real results. It all starts with you, as the emphasis on storytelling and role-modeling in OpenAI’s model highlights.

The Bottom Line: AI Mastery in October 2025

So where does that leave you?

You should feel confident if you can use AI effectively in at least five core workflows (writing, coding, analysis, brainstorming, integration).

If you can also:

• Apply AI with enough fluency to explain and guide others.
• Treat AI not just as a productivity tool, but as a thought partner.
• Shape culture and adoption at the team level.

Then you’re well ahead of the curve.

But waiting for AI to “settle down”?

That could take decades.

Mastery isn’t about waiting for the storm to pass; it’s about learning how to navigate in the storm.

Questions? Thoughts? Share with us here.

- ​Daan