Rolling out Microsoft Copilot or Gemini gives employees access to AI, but this won’t automatically generate organisation-wide productivity gains. While initial excitement around AI is easy to generate, turning it into lasting business value is a very different challenge. With the excitement surrounding AI’s potential to transform the workplace comes the expectation that employees will work faster overnight. Teams will become exponentially more efficient. Costs will plummet, while productivity and ROI soar. It’s an enticing vision that fuels an early wave of AI experimentation. But the challenge for businesses really begins once this initial excitement fades. While executive teams see AI platforms being rolled out and assume adoption is well underway, L&D teams and programme leaders often see a different version of reality. While some employees quickly integrate AI into their everyday work, others struggle to move beyond occasional use. Some wait for clearer guidance, while others develop their own ways of using AI, adopting inconsistent practices that can jeopardise responsible AI use and limit the value realised. In this article, we explain why AI adoption often stalls post-launch and why building an AI-capable workforce is key to maximising a sustainable return on investment. AI adoption takes more than access to a platform Initial curiosity naturally encourages experimentation with new AI tools. But widespread access doesn’t necessarily translate into sustained adoption. According to McKinsey’s State of AI in 2025 report, almost nine in ten organisations are now using AI in at least one business function. But in most cases AI is deployed in isolated pilots and pockets of experimentation, with only around one-third of organisations scaling AI across the enterprise. Sustaining momentum depends on employees moving from occasional use to integrating AI into the tasks they perform every day. Research by law firm Gowling WLG found that while 94% of executives consider AI strategically important, more than half (55%) say scaling AI has been more difficult than expected. Only 7% report strong AI adoption across their organisation. Like any transformational technology, realising AI’s full value requires people to adopt new AI-enhanced ways of working. And this comes with a learning curve. Rather than completing every task themselves, they need to learn how to collaborate with AI tools and agents – how to frame effective prompts, evaluate responses and recognise when human expertise is still needed. This transition takes time, which means productivity improvements often take longer to materialise than leaders expect. The economics of AI adoption In order to embed AI more broadly across departments and workflows, organisations need to extend software access, buy new licences, strengthen cloud infrastructure and increase compute capacity. This introduces additional costs that many organisations underestimate. According to research by WitnessAI, 68% of US organisations experienced AI projects exceeding budget over the past year. Conversely, only 9% reported that 75% or more of their AI initiatives delivered a measurable financial return. When investment outpaces realised value, budgets face greater scrutiny and programme leaders come under increasing pressure to justify further spending. This is when rollout plans lose momentum and AI initiatives deliver only a fraction of their transformative potential. This pattern is reflected in a study by Anthropic. After analysing anonymised AI usage data across more than 800 occupations, researchers found a significant gap between what AI is capable of doing and how it is actually being used at work. For example, although AI has the potential to assist with around 94% of tasks in computer and mathematical occupations, employees currently use it for only around one-third of those activities. Similar gaps exist across business, finance, legal and administrative roles, suggesting the capability of the technology itself is not the limiting factor. Instead, a failure to build workforce capability is driving the AI adoption gap. Closing the AI adoption gap While the narrative that technology should augment human skill rather than replace it is increasingly accepted, there remains a tendency to overlook the critical role of people in AI-assisted processes. AI can automate parts of a task, but employees still need the confidence and judgement to apply it effectively in their day-to-day work. AI-generated code needs testing before it’s released. Financial forecasts need validating before they’re shared with the board. Contracts require legal review. Customer communications must reflect brand standards and regulatory obligations. In every case, AI can accelerate the work, but people remain accountable for the final output. The challenge for employees is learning how to integrate AI into their work effectively. They need to understand where AI performs well and where its limitations lie. They need to ask better questions, evaluate responses critically and recognise when further verification or specialist expertise is required. This is how learning helps sustain momentum after launch, becoming an important enabler of long-term AI adoption. Building workforce capability As Anthropic’s research shows, AI has the potential to support far more workplace tasks than it does today. But unlocking that potential requires investment in building the workforce capability to integrate AI into everyday ways of working. Building this AI capability takes more than a one-hour lunch-and-learn. It develops over time through continuous learning and opportunities to use AI in realistic business scenarios. At Go Courses, we help organisations sustain AI adoption by making it easier to access and deliver high-quality training from leading technology providers, including Google, AWS, VMware and Broadcom. Through role-based learning pathways and our global instructor network, we help organisations align AI learning with business objectives and workforce roles, enabling employees to develop the AI capability needed to use AI confidently and effectively in their day-to-day work. Whether you’re introducing AI across the organisation or looking to maximise the return on your existing AI investment, Go Courses helps you build the workforce capability needed to turn AI potential into lasting business value. Talk to us about scaling AI training
