TECHNOLOGY by Kathy Durfee Built to bend: how flexible leaders handle AI adoption without breaking their business When our team first adopted AI tools, stress quickly replaced curiosity. The relentless upgrades, unfamiliar features, and constant new versions made the transition feel overwhelming, even in tech, where rapid change is the norm. Spotify took over ten years to reach 100 million users. ChatGPT did it in just three months. Suddenly AI was everywhere. On top of this, AI itself keeps evolving rapidly, stacking another layer of complexity for organizations already juggling day-to-day operations. We saw chaotic adoption by our customers. Whether solo founders or hundreds of employees, people were using AI Chat tools to help draft emails, analyze data, summarize documents, respond to customers, and generate ideas—often switching from one tool to another. The adoption process is remarkable not only for its speed but also for how informal and unstructured it often is. Some people approach AI with a do-it-yourself attitude. If they can chat with it, they think they know how to use it well. Podcasts, demos, and social media boost that confidence. To be fair, AI does empower people. It helps each of us move faster, explore options, and overcome hesitation. But using AI without a clear strategy informed by best practices can quietly hurt a business. We’ve noticed that when people use AI in areas they know well, something interesting happens. They start correcting the responses, questioning assumptions, and refining answers. That’s often when AI stops feeling like an authority and is seen for what it is, a powerful support tool. However, not everyone in the company recognizes this shift at the same time. Stage one: the wild west For most organizations, adopting AI at first is like an open frontier where experimentation rules. We’ve seen teams move very quickly in this stage but also make mistakes that could have been avoided. People often assume AI is a knowledge base, that it’s been fully tested, and that it should always be accurate. They think that if they pay for the subscription, their data is safe. In reality, AI is probabilistic software. It gives answers that sound reasonable, but they are not guaranteed to be correct, and even the paid subscriptions don’t necessarily protect data. Using AI without understanding can be dangerous. Like the wild west, there’s a lot of excitement and unease. That tension usually pushes organizations to the next stage. Stage two: awareness without alignment The second stage is recognition. Leaders recognize AI is important and want to adopt it. Some people rely on it a lot, while others avoid it. Multiple tools show up across the company with little coordination. Leaders start asking hard questions. Do we standardize on one platform, like our email or accounting software? Is AI different? What does “good” adoption look like? We’ve seen organizations struggle at this stage. There’s a desire to innovate but also growing discomfort about inconsistency. They want to move fast and innovate, while also reducing risk. That discomfort often motivates leaders to move from observation to action. Stage three: guardrails and core skills In the third stage, rising concern leads organizations to establish guardrails and build foundational AI skills. Organizations start putting guardrails in place to protect their business. This is when AI usage policies begin to appear. Leaders take stock of which tools are being used and where. They start focusing on data exposure, intellectual property, and accuracy. In practice, this stage often reveals gaps. Traditional data loss prevention tools typically don’t address AI-related challenges, and technical controls alone aren’t enough. The human side becomes just as important. Do people know how to ask good questions? Can they tell when an answer sounds confident but is not correct? Are they comfortable checking information before acting on it? Do they know which tools are approved and which are not? We have seen cases across industries, from manufacturing to finance, where even though the company has an approved list of tools and an AI usage policy, individuals use personal, paid-for AI tools, confident they are helping, not hurting, their organization. wacomka / Shutterstock.com 92 enterprising Women
RkJQdWJsaXNoZXIy MzA5OTEy