How Executives Are Actually Using AI at Work (Not How They Say They Are)
There's what executives say about AI in public. And there's what they actually do with it behind closed doors.
In public: "We're exploring strategic AI initiatives across the organisation."
In private: "I used Claude to rewrite my board paper last night and it was better than what I'd have produced in three times the time."
I've had this conversation with over 50 executives in the last year. CEOs, CFOs, board members, managing directors. Across resources, professional services, infrastructure, and trades businesses in Australia.
The pattern is clear. The executives who are getting real value from AI aren't talking about it in strategy documents. They're using it quietly, personally, and practically.
Here's what they're actually doing.
The Quiet Adoption
Most executive AI adoption looks the same.
It starts with curiosity. Someone mentions it at a dinner. A board member asks about it. A direct report starts producing unusually good work suspiciously fast.
Then there's the experiment. Usually ChatGPT. Usually a simple question. Usually disappointing.
Then — for about 30% of executives I talk to — there's the second attempt. This time with more context. A longer prompt. A real task. And it works.
That second attempt is the inflection point. The executives who push past the initial disappointment discover something transformative. The ones who don't go back to doing things the old way.
The difference between the two groups isn't intelligence or technical ability. It's willingness to invest 10 hours learning how to prompt properly. That's it.
Use Case 1: Board and Committee Paper Preparation
This is the most common use case by far. And the highest-value one.
Every executive I know who has adopted AI for board work describes the same experience:
Before: 3–4 hours per paper. Staring at a blank page. Writing, rewriting. Formatting. Checking.
After: 30–45 minutes. Feed the AI the source material, the structure requirements, and the institutional context. Get back a draft. Edit for judgement and politics. Done.
One CFO I work with told me: "I produce twice as many board papers now. Not because anyone asked me to. Because I can. And the quality of decision-making in the boardroom has improved because directors have better information."
That's the compound effect. Better tools don't just save time. They change what's possible.
The key technique: system prompts. A reusable instruction set that tells the AI exactly how your board papers should be structured, what tone to use, what format to follow. I've written about this in detail in my board paper preparation guide.
Use Case 2: Deal Analysis and Due Diligence
For executives involved in M&A, capital raising, or major commercial decisions, AI has become a critical analysis tool.
What I see in practice:
Information memorandum review. A 200-page IM used to take a full day to read and annotate. Now: feed it to the AI, ask for a summary of key risks, financial highlights, and inconsistencies. First-pass analysis in 15 minutes.
Financial model review. "Check this DCF model for logical errors and flag any assumptions that look aggressive compared to industry benchmarks." Not a replacement for your financial adviser. But a brutally effective first screen.
Contract comparison. "Compare these two lease agreements and highlight the key differences in terms, obligations, and exit provisions." What used to require a junior lawyer and 4 hours now takes 5 minutes for first-pass analysis.
Competitor analysis. Feed in three annual reports and ask for a comparative analysis across revenue growth, margins, capex intensity, and strategic direction. Synthesis that would take a half-day done in 10 minutes.
A managing director of a mid-market PE fund told me: "We're not doing more deals. We're doing better analysis on every deal. Our hit rate has improved because we're catching issues earlier in the process."
Use Case 3: Communication and Stakeholder Management
This one surprised me. I expected AI to be primarily about analysis. But the executives getting the most value are using it for communication.
Investor updates. Quarterly investor letters that used to take a full day: now 90 minutes. The AI handles the structure and data narrative. The executive adds the strategic commentary and judgement calls.
Team communications. Restructuring announcements, strategy updates, change management messaging. AI produces the first draft. The executive refines tone and nuance.
Presentation preparation. Not just slides — but the talking points, the anticipated questions, and the response frameworks. One CEO told me she feeds the AI her board pack before every meeting and asks: "What are the five hardest questions I'm likely to face, and what's the best answer to each?"
Email triage. Senior executives receive 150–300 emails per day. AI can draft responses, prioritise by urgency and strategic importance, and flag items that need personal attention.
Use Case 4: Strategic Thinking and Decision Frameworks
This is the most sophisticated use case and the one with the highest potential value.
Scenario planning. "Given these three market scenarios, what are the implications for our business and what should we do in each case?" The AI won't give you the answer. But it'll give you a structured framework for thinking through the scenarios.
Decision documentation. Before making a major decision, feed the AI the context and ask it to construct the strongest argument FOR the decision and the strongest argument AGAINST. Having both arguments articulated clearly is more valuable than any amount of confirmation bias.
Risk identification. "Here's our strategic plan. What are we missing? What could go wrong that we haven't considered?" AI is excellent at identifying blind spots because it doesn't share your biases.
Analogy and pattern matching. "What industries have gone through a similar transition, and what happened?" AI's breadth of knowledge makes it good at finding parallels that humans — trapped in their industry silo — would miss.
What the Non-Adopters Are Missing
The executives who aren't using AI aren't losing anything today. Their board papers are still fine. Their deal analysis still works. They're still effective.
But the gap is widening.
The AI-adopting executives are producing more, higher-quality work. They're better prepared for meetings. They're making better-informed decisions. They're spending less time on production and more time on judgement.
In six months, that gap will be visible. In a year, it will be significant. In two years, not using AI as an executive will be like not using email in 2005 — technically possible, but a clear competitive disadvantage.
The Barrier Isn't Technology
Every executive has access to ChatGPT or Claude. The tools cost less than a business lunch.
The barrier is skill. Knowing how to prompt effectively. Understanding what AI does well and what it doesn't. Building workflows that integrate AI into your existing processes rather than adding it as a separate step.
This isn't a technology adoption problem. It's a skill development problem.
And like most skill development, it's faster with structured guidance than with trial and error.
The Privacy and Confidentiality Question
Every executive I talk to raises this. Rightly.
"Can I put confidential board material into an AI?"
The answer depends on which tool and which plan you're using.
Enterprise plans (Claude for Enterprise, ChatGPT Enterprise) don't use your data for training. Your inputs are not shared. This is contractually guaranteed.
Consumer plans vary. Read the terms of service. When in doubt, use the enterprise tier.
Practical approach: For genuinely sensitive material (M&A activity before announcement, price-sensitive information, personnel matters), I use an enterprise-grade tool with appropriate data handling policies. For general analysis and non-sensitive work, the standard plans are fine.
Don't let the confidentiality question stop you from starting. Use non-sensitive material first. Build the skills. Then upgrade to enterprise when you're ready to use AI for your most sensitive work.
Getting Started This Week
You don't need to transform your workflow overnight.
Today: Sign up for Claude or ChatGPT if you haven't already. The free tier works for learning.
This week: Take one document you're working on — a memo, a report, an email — and draft it with AI. Feed the AI your brief and source material. See what comes back.
This month: Build your first system prompt for the document type you produce most often. Test it on three real documents.
Within 90 days: You'll wonder how you worked without it.
That's not optimism. That's the consistent experience of every executive I've worked with who committed to learning this properly.
The gains are visceral. You feel them.
I teach executives the practical AI skills that actually matter — system prompts, workflows, board paper prep, deal analysis. Join the next cohort — 4 weeks, step-by-step, free.