It’s no exaggeration to say that artificial intelligence (AI) is the most significant advancement — technical or otherwise — in our lifetimes. AI, like steam before it, is powering the Fourth Industrial Revolution. As Seth Godin has said, AI is the biggest change to our world since electricity. And because AI usage is on track to be as ubiquitous as Googling something is today, the technology is likely to spark not just significant but exponential economic growth.

As AI rapidly becomes something that nearly anyone can use, no business leader has to be convinced that the world is changing. As a recent McKinsey report put it, exploring generative AI is not a maybe for companies — but a must.

Today, all companies are essentially tech companies, which means they’ll soon have to modify their digital processes to accommodate AI. That means leaders across the enterprise need training to learn more about AI. However, research shows that most corporate leaders have never attended an AI training. So, what exactly do they need to know?

Here are four things business leaders should know about AI so they can focus their training efforts around AI — and better understand what this fast-changing technology means for their companies, their clients and their customers.

AI: 4 Things Your Leaders Need to Know

1.    AI is not software.

The great thing about software is that it either works or it doesn’t. When someone fills out and submits an online form, for example, it sends the data unless there’s a bug.

AI, on the other hand, is not software. It works in degrees of accuracy, which means it’s not correct 100% of the time. Sometimes ChatGPT won’t produce the most polished or accurate piece of writing. Sometimes that self-driving car will run into something. Even when AI is incorrect, it will continue to function.

This requires company leaders to understand that AI can get things wrong — and to be prepared to deal with the consequences when it inevitably happens. However, only 42% of corporate leaders report feeling very or extremely confident they can use AI tools without compromising company data today, and more than one-quarter are not very or not at all confident.

Remember: It’s important to keep in mind that AI essentially is simply a feature or a new type of user interface built into a broader program. Its true value to an organization comes from its integration into other applications.

Whether company leaders see AI integration into their businesses as incremental or transformational, they should work to understand the technology’s potential — and its limitations — so they can evaluate AI opportunities across an organization.

2.    AI ethics and policies are critical.

Every company should adopt ethical policies and principles around a set of themes: transparency, explainability, human agency and oversight, accountability and compliance, technical robustness and safety, privacy and data governance, diversity and nondiscrimination, and societal and environmental impact. These policies should guide how companies deploy AI within their products and services and how AI interacts with customers and clients.

It’s not just a matter of doing the right thing. It’s a growing legal imperative as well. The EU Artificial Intelligence Act — the world’s first major comprehensive AI legal framework — puts the onus of responsible AI use on providers and developers, who must comply with numerous regulations in order to do business in the EU. Regulation in the U.S. is more piecemeal. No federal statutes currently govern AI, but 27 states have enacted laws directly or tangentially related to AI. That suggests a complicated legal landscape that could grow increasingly complex.

Because the questions surrounding AI are so large and multidimensional, organizations should consider hiring or contracting for the services of an AI ethicist, especially when issues of potential societal bias arise.

3.    AI training must extend throughout the entire organization.

Talent managers are increasingly seeking new hires with AI skills even for roles not focused on AI. That suggests that everyone in an organization needs AI competencies to be successful.

Even more immediately, legal, compliance and procurement teams should receive AI training so they can understand not only the shifting regulatory landscape but also the rapid growth of an entirely new tech ecosystem that has foundational AI models at its core. As organizations mull build-versus-buy decisions, they’ll have to be able to evaluate vendors and the AI opportunities they’re providing. Today, nearly 2 in 5 leaders (39%) say they don’t feel they have the knowledge to make decisions about AI vendors.

To further understand AI and enhance training opportunities, organizations should establish and empower a cross-functional team with representatives from the product, design, data science, engineering and legal departments. This team should focus on the opportunities and risks of using AI and coordinate AI strategy throughout the organization. A multi-disciplinary team approach tears down the silos that have led to so many misconceptions and misunderstandings of AI.

AI requires everyone’s involvement, but at the least should start with compliance, procurement and product management. A company-wide team that’s leading AI conversations can identify critical issues and trends, get ahead of challenges, direct training efforts and help an organization leverage AI to its fullest potential.

4.    A growth mindset is required.

AI is complicated. It’s OK not to understand it — and it’s OK to admit that.

Company leaders should adopt a growth mindset when it comes to AI. Harnessing AI’s power takes collaboration and innovation. It requires input from throughout an organization to devise sound strategies around AI. To learn and understand AI, organizations should bring in subject matter experts (SMEs) who know what’s happening now, how it’s changing and what challenges await just over the horizon.

In the early days of the World Wide Web, changing a website required help from a computer programmer. Eventually, content management systems took control from data scientists and put it in the hands of everyday users.

Today, anyone can build a website. Because of emerging applications such as ChatGPT and Dall-E, we’re on the cusp of a similar shift in AI. Soon, everyone — not just computer engineers — will be able to use AI across nearly every facet of the workplace.

As access becomes broader, new technologies, new opportunities and exponential growth will follow. Today, only 16% of corporate leaders report that their companies provide regular AI training to employees. But achieving an AI-confident workforce, and the competitive advantage that comes with it, will require a much greater investment in training and policy development, which will need to come from the top.