A 2025 McKinsey report found that business leaders feel that employees are frustrating their plans to capitalize on AI-driven efficiencies. The report says that C-suite executives are 2.4 times more likely to cite employee readiness as a barrier to AI adoption than they are to point to leadership alignment issues.
A recent event in the world of artificial intelligence suggests employees are not shy at all about using AI in the workplace, however. In fact, the major ChatGPT outage that occurred on June 10, 2025, revealed employees have come to rely on AI to help them with their weekly workloads, driving automation and other efficiency boosts even in environments where AI use is discouraged.
“Looking at this outage data, what really jumps out is how the corporate world’s relationship with AI tools has evolved into something quite different from what IT departments imagine,” says Dev Nag, Founder and CEO of QueryPal, an enterprise-grade agentic AI solution for customer experience. “The 400 million weekly users experiencing disruption weren’t primarily using expensive enterprise licenses or carefully vetted corporate deployments. Around 70 percent were accessing free, consumer-grade ChatGPT through their browsers, often without their company’s knowledge or blessing.”
Nag’s insights suggest that the easy accessibility of generative AI tools and the assistance they can provide to employees at all levels has accelerated AI adoption to a degree many business leaders don’t fully understand. Although AI integration strategies are being debated in high-level meetings, employees are already using these tools to add efficiency to their daily tasks, creating a new phenomenon known as shadow AI.
“This shadow usage has created a fascinating dynamic where the real AI transformation in workplaces is happening through the back door, driven by individual workers who discovered these tools make their jobs easier and just started using them,” Nag explains.
What Happened When ChatGPT Went Down?
On June 10, 2025, ChatGPT experienced an outage lasting more than a day. Workers who have integrated the generative AI system into their workflows quickly found alternatives, even if they offered less functionality.
AI on PTO: The June 2025 ChatGPT Outage
“Massive” was the term media outlets chose to describe the ChatGPT outage that occurred on June 10 at approximately 8:00 a.m. British Summer Time as workers in the UK were making their way to the office. At the time, OpenAI said they were investigating reports of widespread failures.
The outage, which impacted free and paid-for tiers of ChatGPT as well as OpenAI’s Sora image generator, sent users scrambling for information. Google searches on the status of OpenAI platforms became the second most popular search that day, prompting more than 500,000 queries.
Many of those looking for info were panicked, especially those who had come to depend on the AI platform for work task automation. As shared in an Inc. article covering the outage, one X user posted, “Suddenly ChatGPT goes down and I no longer know how to work.”
“When ChatGPT went down for 10 hours, it wasn’t the engineering departments that panicked; it was the people who’d quietly integrated AI into dozens of daily tasks without ever filing an IT request,” Nag says. “The 500,000 Google searches came largely from workers who’d built their entire workflow around a tool their company didn't even know they were using.”
What the Outage Means for the Future of Workplace AI
AI-driven tools, from an enterprise perspective, are in their early days. Companies in the process of bringing them onboard are still determining what features are most important and what challenges they need to consider.
The ChatGPT outage suggests that both companies preparing for AI integration and those developing AI-driven tools should prioritize reliability over all other considerations. As the response to the outage revealed, users weren’t choosy about features when their first option was suddenly unavailable. If companies ultimately embrace that attitude as their employees have, it could disrupt the dynamics in the current tech market.
“The competitive dynamics exposed by the outage challenge conventional wisdom about market dominance in tech,” explains Nag. “Unlike traditional enterprise software, where switching costs create multi-year lock-ins, users demonstrated they could jump between ChatGPT, Claude, Grok or Gemini with minimal friction and in a single prompt. The spike in searches for competing tools during the outage wasn't just desperation; many users permanently added backup AI tools to their workflow that day. This fluidity means reliability has become more valuable than having the absolute best model.”
Developers focusing their energy on perfecting bells and whistles must be careful that added features don’t degrade the platform’s reliability. The lesson to be learned from the ChatGPT outage is that serving users rather than attracting buyers is key to emerging as a leading choice. Algorithmic bias and other ethical concerns must be addressed, of course, but a platform that users can’t access reliably won’t succeed regardless of how trustworthy its responses are.
“A slightly less capable AI that stays online 99.9 percent of the time will win more users than a cutting-edge model that goes down during critical work hours,” Nag adds. “Companies are learning that in the AI productivity market, being consistently available matters more than being occasionally brilliant.”
The outage also put companies on notice that their security frameworks must address shadow AI usage. Many employees use the free, public versions of generative AI platforms even when company policies prohibit it. And to make matters worse, those employees step into an IT role — seeking solutions for outages — when those platforms fail.
Security teams can certainly imagine how an employee would have responded to a phishing attack promising access to ChatGPT at any point during the 34 hours the platform was down. Teams that aren’t addressing that type of risk ignore a major security vulnerability.
Shine a Light on Shadow AI
Technology advancements over the past few years have set the stage for an AI revolution. Costs are down and capabilities are up, inspiring 94 percent of businesses to increase their investments in AI in 2025. And many of the key issues causing concern, such as algorithmic bias, are becoming better understood and more effectively addressed.
Companies must acknowledge that a foundation has already been laid to maximize the ROI on AI, however. The ChatGPT outage has brought shadow AI use into the open, allowing companies to understand and build effectively on the processes their employees have already embraced.
