We've spent the last six months watching enterprise leaders make AI decisions. The pattern is clear: the quality of an organisation's AI strategy correlates directly with how well its leadership understands the basics. Not the technical details. The basics.
What You Need to Know
- AI literacy is not technical expertise. It's the ability to understand what AI can do, what it can't do, where it creates value, and where it creates risk. Leaders need this to make informed strategic decisions.
- The three literacies framework we developed in 2020 (health, financial, technology) applies directly to the AI context. AI literacy is an extension of the third literacy: technology understanding beyond your specialism.
- Leaders without AI literacy default to one of two failure modes: uncritical enthusiasm (approving everything) or defensive scepticism (blocking everything). Both are expensive.
- AI literacy can be built in weeks, not months. The goal isn't to make leaders into engineers. It's to give them enough understanding to ask the right questions and evaluate the answers.
44%
of senior executives say they can't evaluate AI vendor claims
Source: Deloitte, State of AI in the Enterprise, 5th Edition, 2022
The Literacy Gap
Here's a conversation I had last month with a CEO. It's representative of dozens like it.
"We need an AI strategy. Our competitors are doing AI. The board wants to see progress. Can you help us pick the right model?"
That question - "help us pick the right model" - tells me everything I need to know about where this organisation is. They've skipped past the strategic questions (what problem are we solving? what does our data look like? where does AI create measurable value?) and jumped straight to a technology decision that should come much later.
This isn't a criticism. It's a literacy gap. The CEO is smart, experienced, and motivated. They just don't have enough context about how AI works to know which questions to ask first.
The same gap plays out across the C-suite:
- CFOs who can't evaluate whether an AI investment will deliver returns because they don't understand the difference between a proof of concept and a production system
- COOs who approve AI pilots without understanding the integration and change management costs
- CTOs who focus on model selection when the real constraint is data quality
- CROs who sign AI vendor contracts without knowing what questions to ask about accuracy, governance, or lock-in
The Three Literacies, Extended
In 2020, we published The Three Literacies: health literacy, financial literacy, and technology literacy as interconnected professional capabilities. The idea was simple - people who understand more of the world they operate in make better decisions.
AI has made the third literacy (technology literacy) dramatically more urgent.
When we wrote that piece, technology literacy for leaders meant understanding how software gets built, how digital products work, and how technology creates business value. That baseline still applies. But AI adds specific dimensions that leaders need to grasp:
Understanding AI Capability and Limitation
Leaders need a practical mental model of what current AI can do well (pattern recognition, language processing, structured analysis) and what it can't do well (genuine reasoning about novel situations, reliable factual recall, understanding of context it wasn't trained on).
This doesn't mean understanding transformer architectures. It means understanding that an AI can summarise your policy documents accurately but shouldn't be trusted to make coverage decisions without human review. It means understanding that "95% accurate" still means 1 in 20 outputs needs correction.
Understanding Data as a Strategic Asset
AI capability is bounded by data quality. Every enterprise AI conversation eventually arrives at the same place: "our data isn't ready." Leaders who understand this upfront invest in data readiness before AI implementation. Leaders who don't discover it three months and $200K into a pilot.
Understanding the Build vs Buy Spectrum
AI is not a single purchase decision. It's a spectrum of options: use an off-the-shelf tool, configure a platform, build a custom solution, or (most commonly) a combination. Leaders who understand the trade-offs make better investment decisions. Leaders who don't get sold whatever the first vendor pitches.
Understanding AI Risk
AI introduces specific risks that differ from traditional technology: bias, hallucination, privacy exposure, vendor dependency, regulatory uncertainty. Leaders need enough literacy to ask about these risks and evaluate the answers, not enough to solve them personally.
Building AI Literacy in Practice
The good news is that AI literacy for leaders can be built quickly. We're not talking about a six-month programme. We're talking about focused learning that changes the quality of strategic conversations.
Week 1-2: Foundations. What AI is and isn't. How large language models work (conceptually, not technically). What "training data" means. What hallucination is and why it happens. Hands-on time with ChatGPT or similar tools.
Week 3-4: Enterprise context. How AI applies to your specific industry. What your competitors and peers are doing. What good AI governance looks like. How to evaluate vendor claims. What questions to ask.
Week 5-6: Strategic application. Where AI creates value in your operations. What your data readiness looks like. What investment is required and over what timeframe. How to measure AI outcomes.
After six weeks, a leadership team won't be AI experts. But they'll be able to have informed conversations about AI strategy, ask the right questions when evaluating proposals, and make decisions grounded in understanding rather than hype or fear.
The Litmus Test
After an AI literacy programme, a leader should be able to answer: "If someone proposes an AI initiative, what are the first five questions I should ask?" If they can't, the programme hasn't worked.
The Cost of Illiteracy
We've seen the cost of AI illiteracy firsthand this year. Organisations approving $500K AI investments based on vendor demos. Boards demanding "AI strategies" without understanding what that entails. Leadership teams making technology decisions that should be business decisions, and business decisions that should be technology decisions.
The enterprises that will lead in AI aren't necessarily the ones with the biggest budgets or the best engineers. They're the ones whose leadership understands enough about AI to make informed decisions about where to invest, what to build, and how to govern it.
That understanding starts with literacy. And the best time to start building it was six months ago. The second best time is now.