Senior leaders from Mastercard and AI Truth on fostering shared understanding – so teams know when to trust or challenge AI, and how to integrate it into a process
By Eoin Connolly
Over the last few years, AI has expanded beyond being the sole preserve of data scientists. Tools like ChatGPT, decision intelligence systems, and even early-stage agentic AI platforms are now in the hands of employees across every department. This radical shift in the modern business environment has come as a response to AI’s game-changing potential.
But the more employees that use AI in their day-to-day duties, the greater the risk that a lack of education could cause critical problems. Education, though, isn’t a straightforward topic in itself. Running awareness sessions isn’t going to cut it: true AI education needs to be practical, contextual, and laid out in as structured a manner as possible.
JoAnn Stonier, Fellow of Data and AI at Mastercard, and Cortnie Abercrombie, founder of AI Truth, have spent years working on how organizations can use AI responsibly. Their insights reveal where most corporate AI education programs go wrong, and how today’s organizations can do better.
Education without application falls flat
For Stonier, the most common pitfall is treating AI awareness as a standalone compliance exercise. “Just because you’ve told people what AI is, doesn’t mean they understand how it applies to their work, or how to use it safely. Without context, awareness doesn’t change behavior.”
Abercrombie echoes the point, adding that generic awareness training often fails to stick. “People need to see how AI affects their role, their deliverables, their decision-making. Otherwise it’s just background noise.”
Both argue for role-specific education that bridges the gap between technical capability and business need. Making this a priority won’t just help individual employees with their work. It’ll also contribute to the broader, inter-departmental cultural shift that so many businesses are prioritizing more and more all the time.
Personal relevance drives retention
Abercrombie stresses that connecting education to personal stakes, not just corporate policy, can make the message resonate. “If you want people to remember, make it personal,” she says. “Explain what it means for their reputation, their ability to do their job well, and their career progression.”
Stonier has found success with a similar approach, reframing AI risk in relatable terms.
“If I ask someone, ‘Would you put your bank account number into a chatbot?’ they immediately see the danger,” she explains. “Now they can connect that same logic to protecting the company’s data.”
It’s an intelligent way to approach the problem: bringing abstract, high-level concerns down to the practical ins-and-outs of daily work. But what’s the best way to make that ideal a practical reality?
For both leaders, education is inseparable from governance. After all, employees can only follow the rules if they understand why they exist and how they apply. Throwing an enormous policy document at your workers and hoping they’ll read it cover-to-cover just isn’t going to cut it.
“You can’t rely on policy alone,” explains Stonier. “You have to make sure people know what’s acceptable, what’s not, and why. Otherwise you’re asking them to navigate a black box.”
“Governance sets the boundaries, but education gives people the skills and confidence to work inside them,” echoes Abercrombie, who recommends integrating education into governance rollouts so that every policy comes with practical guidance and real-world examples.
The lack of AI literacy creates misalignment between business and technical teams. And the consequences are significant: slower adoption, lower ROI, and an ever-increasing chance your competitors might get the jump on you.
As Abercrombie puts it: “If business leaders can’t articulate what they need from AI, and data teams can’t explain what’s possible, you get wasted effort and lost opportunity. We need to teach people to think critically about AI outputs, not blindly accept them. That means asking: does this align with my business goal? Does it meet compliance requirements? Does it make sense?”
The literacy gap won’t be solved by one-off workshops, though. It requires ongoing dialogue between functions, supported by shared language and metrics.
“Without shared understanding, you get AI projects that are technically brilliant but strategically irrelevant,” says Stonier. “AI literacy isn’t just knowing what a model does. It’s knowing when to trust it, when to challenge it, and how to integrate it into a process.”
Training for the AI we haven’t met yet
The pace of AI change means education programs can’t just focus on today’s tools, exacerbating the challenges of effectively educating workforces across sectors. Adaptability is going to be the name of the game moving forward, especially given the blistering pace of AI development.
“If your training is tied to a single platform, it’s obsolete as soon as the next one comes along,” says Abercrombie. “You need to build principles that apply no matter what tool they’re using.”
Stonier agrees: “The AI they’re using in three years will look nothing like what they’re using today. So the core skills – critical thinking, risk awareness, ethical reasoning – those are what last. Make AI literacy part of onboarding, part of project kick-offs, part of performance reviews. It should be woven into how the organization operates.”
Raising AI literacy isn’t about checking a box on an awareness program. It’s about building a culture where employees are both confident and cautious in their use of AI, and where they understand their individual role in the broader organizational ecosystem. Cross-departmental collaboration and practically-minded educational initiatives are pieces of the same puzzle. Solving it could unlock major ROI in the years to come.
“Education alone isn’t enough,” Stonier says. “You have to connect it to practice, to governance, and to people’s real work. That’s when it changes behavior.”