Major technology waves usually follow the same arc. There’s a quick, obvious, and contagious upside early. Then the problems follow.

The internet gave businesses global reach, fast communication and previously unimaginable access to information – along with data breaches and security risks that nobody could have anticipated. Organizations scrambled to catch up with firewalls, acceptable use policies and IT oversight.

Generative AI is following a similar pattern. Adoption has been swift and largely uncoordinated. Inappropriate and unsecure use can lead to more problems than the value it creates.Only 22% of CEOs reporthaving a comprehensiveAI governanceplan in place. So, whilethe productivity gainsare real,governance is not the realityfor most small and medium-sized businesses.

That gap is a problem that exposes organizations to tremendous risks. It also turns out that closing this gap can unlock something more valuable than most CEOs realize.

VISTAGE CEO CLIMB ENearly Half of CEOs Are Flying Without a Net

In addition to not even a quarter of CEOs having an AI governance plan, the same Vistage research shows another 21% have only an informal or partial policy — which amounts to no meaningful plan — 26% are actively developing one, and 27% have no policy and no plans to create one.

In other words, many CEOs are running organizations without adequate AI guardrails. These risks aren’t theoretical. In 2023, Samsung engineers inadvertently exposed sensitive source code to ChatGPT. Although the disclosure appears to have been unintentional, once confidential information is submitted to a public large language model, it can be effectively lost and, even more dangerous, potentially reproduced or exposed. Clear, well-defined protocols are precisely what prevent incidents like this.

A Trio of Pillars, A Single Priority

For those CEOs looking for a governance plan, research points to 3 core components as critical:

1. Acceptable usesets the standards for how AI can and cannot be used in a professional space. Among CEOs with a governance plan,87% have these guidelines in place. It’s the equivalent of an internet use policy for the AI era. It establishes appropriate behaviors and a framework for enforcement.

2. Data securitygoverns what can be fed into AI tools. Anything entered into an unsecured model can become part of its training data and, potentially, become public. Eighty-six percent of CEOs with a governance plan have these protections in place. Proprietary information and client data should never go into a public AI tool.

3. While these pillars protect, the third competency,approvedtools, is where opportunity truly lies. Only 71% of CEOs with a governance plan have standardized or approved tools, the lowest adoption rate of the three pillars — and the most consequential gap. Without a common toolset, even organizations that get security and acceptable use correct risk leaving value on the table.

The Tower of Babel or a Collaborative Common Platform?

Here’s what’s actually happening in most businesses today: Individual employees are finding ways to use AI on their own and in their own way, building workflows nobody else can see or replicate.The gains are realbut random.

This is the Tower of Babel problem. When a sales representative is on ChatGPT, a finance manager is on Copilot, and a marketing director is on Claude, they’re each getting individual value — but they can’t share it. The tools largely do the same thing, but with different terminologies, structures, approaches, and best practices. The workflows don’t translate. Everyone is learning different things at different paces, and the organization doesn’t get smarter.

The dynamic is changed once there’s a common set of approved tools. When everyone is working on the same platform, teams can share best practices, use the same terminology, and work in the same shared structures. Individual productivity goes from random spots of value to an organization that optimizes overall impact. In essence, the entire organization lifts itself up.

That common platform for collaboration is the first key to organizational optimization.

The CEO’s Move

A couple of deliberate decisions will set CEOs on the right path. First, inventory what’s already running in the organization. Employees are using AI tools right now— some sanctioned, some not. Learn about the landscape before standardizing it. The tool sprawl is almost certainly bigger than it appears.

Next, define the approved toolset and commit to it. Communicate the decision clearly and explain why it matters. The goal is to channel AI energy.

Vistage Researchprojects that more than half of CEOs will have comprehensive governance plans in place by next year. The organizations building that foundation now won’t just be better protected; they’ll also have a head start on the collaboration advantage that tool-sprawled competitors can’t replicate quickly. That’s how you power up an organization in this first era of AI.

This story first appeared inInc.