Why companies must move from AI usage to AI upskilling
By Niv Liran, Chief Product & Technology Officer at Unzer
Artificial intelligence is quickly shifting from testing phases to daily use and from chatbots to agentic work. In the past two years alone, tools that were once used mainly by data scientists and engineers have become available to employees throughout companies. Generative AI, in particular, has made these tools far easier to use. Today, anyone can use powerful AI tools with natural language.
But while the availability of AI tools has grown quickly, regular and effective use has not always followed as quickly. Many companies are seeing the same pattern: employees try AI a few times, initial interest is high, and then usage levels off without changing everyday work in a meaningful way.
This gap shows a key point: AI adoption is not mainly about rolling out tools. It is about teaching employees how to use AI effectively throughout the company. Just as basic digital skills became essential when the internet first spread in workplaces, AI literacy is quickly becoming a core professional skill.
From access to capability
When organizations roll out new technology, the instinct is often to make sure everyone can use it. If employees have access to the tools, the thinking goes, they will start using them on their own and output will increase.
With AI, however, access by itself rarely changes how people work. As there is a vast amount of potential use-cases, employees might test a tool, but without a clear understanding of its potential, what can it do, when does it work reliably, and how it fits into current work processes, people only use it in limited ways.
Real change happens when employees start to see AI not as a new gadget but as a useful tool in their day-to-day work. That shift requires trust in the tool, and trust comes from practice and familiarity. This is why organizations now need to go further than simply giving employees tools and teach employees how to use them effectively. Employees need the time to learn and practice so they can understand clearly how AI can help with their actual job tasks.
AI adoption is function-led, not industry-led
One of the clearest trends appearing in many industries is that AI adoption is not mainly determined by the sector a company works in. Instead, it is largely determined by the kind of work people do. Roles that involve writing, analyzing information, coding, designing, or creating content have started using AI faster than others. These information-heavy tasks match well with what generative AI systems can do, which is work with language, produce text, and organize information.
Operational and frontline roles, in contrast, often have fewer direct ways to use AI tools. This does not mean these roles are less able to benefit from AI. The challenge is that the technology is not integrated directly into operational tools, software and processes.
If organizations ignore this gap, there is a chance that efficiency improvements from AI will appear mainly in certain functions while others gain little. Over time, this may lead to larger differences in productivity and career opportunities between roles inside the same company. For that reason, one of the most urgent steps in AI adoption is embedding AI into core operational tools and systems so that its advantages are not limited to office work.
Understanding what AI does well, and where it struggles
Another key part of AI literacy is knowing both the strengths and limits of the technology itself. Generative AI is already very good at short, clearly defined tasks. Drafting text, summarizing documents, translating information, structuring ideas, or generating code are all areas where AI can help people complete work faster. In many cases, these tools can finish these tasks faster than a person could or produce a useful first draft that employees can improve.
At the same time, AI systems still perform poorly with tasks that depend on large amounts of context, careful judgment, fully deterministic outcome or many coordinated steps. Long projects that involve specialized expertise, long-term planning, or unclear decision situations remain areas where human expertise is essential.
Understanding this distinction is critical. When employees know where AI performs well, they can apply it effectively to the right problems (And avoid spending time on the wrong ones). When they understand its limitations, they can apply appropriate oversight and avoid overreliance. In practice, this means AI is currently most powerful as a tool removing repetitive or time-consuming tasks, allowing employees to focus more on analysis, decision-making, creativity, and interpersonal collaboration.
The impact of training on productivity
Perhaps the most important finding from recent studies and discussions among companies is the importance of training in shaping the results of AI use. Studies repeatedly show that employees who receive training in AI tools achieve much larger productivity improvements than those who figure out the tools on their own. In some cases, trained employees are about twice as productive when using AI compared with colleagues who have not received formal instruction.
The difference is substantial. Employees who understand how to apply AI effectively often save many hours per week by automating repetitive tasks, accelerating research, or improving the quality of their outputs. Interestingly, these productivity differences are not primarily linked to age or generational familiarity with digital technology. In many cases, older employees who receive structured AI training outperform younger colleagues who might be assumed to be more comfortable with digital tools but have not received guidance.
This finding reinforces a critical point: AI capability is not about being a “digital native.” It is about learning how to work with new technology in a thoughtful and structured way.
What AI literacy really means
When people hear the term AI literacy, they sometimes assume it refers to deep technical knowledge in machine learning or data science. In reality, the concept is far simpler and more practical. AI literacy means knowing the basics of how AI systems work and how they should be used at work. It includes knowing what types of problems AI can help solve, recognizing its limitations, and developing the ability to critically evaluate the outputs these systems generate.
Equally important is understanding the risks associated with AI. Issues such as data privacy, bias, reliability, and security require careful consideration. Employees need to know when AI tools are appropriate and when they should not be used, as well as what information to share with AI and what not. In essence, AI literacy is about enabling employees to work confidently and responsibly with AI as part of their everyday professional toolkit.
Building a culture of experimentation
Technology changes rarely succeed through top-down programs alone. While leadership can set priorities and provide tools, real change happens when employees start testing tools themselves inside their teams and workflows. Creating this environment requires a workplace culture where people feel encouraged to explore and safely ask questions. Employees need to feel comfortable exploring new tools, testing ideas, and sharing both successes and lessons learned.
One effective way to make this easier is by creating groups of internal AI champions. These individuals connect central technology teams and different departments. Because they know the real challenges and workflows within their teams, they can help turn general AI features into specific practical applications.
In many organizations, some of the most innovative applications of AI are emerging from unexpected places. Sales teams are using AI to prepare customer briefings more efficiently. Compliance teams are exploring ways to analyze large volumes of documents. Operations teams are experimenting with AI-supported workflows to improve response times.
These examples demonstrate that innovation does not need to be confined to technical departments. When employees across functions feel empowered to experiment, new possibilities emerge organically.
Responsible AI as a foundation
As AI becomes used more widely in day-to-day business work, using it responsibly becomes more important. This is particularly true in industries where trust, compliance, and regulatory oversight play a central role. Organizations must create clear rules and oversight structures to ensure that AI is used carefully and appropriately. Data protection, transparency, and human oversight should remain central principles.
AI systems can assist with analysis and decision-making, but accountability must always remain with people. Employees need to understand not only how to use AI effectively but also where the boundaries lie.
Responsible AI is therefore not a separate topic from AI literacy. It is an essential component of it.
The next evolution of workplace skills
If we look back at earlier technology shifts, we see a similar pattern. When the computers were introduced, , computer skills (And learning them) were mostly limited to specialists. Internet was initially a thing for computer geeks. Over time, however, basic digital skills became a standard requirement for most professional roles. The same transition is now happening with AI.
In the coming years, AI literacy will likely become as fundamental as digital literacy is today. Employees across departments will be expected to understand how to use AI tools appropriately, how to interpret their outputs, and how to integrate them into everyday work.
Organizations that succeed in this transition will not necessarily be those with the most advanced technology. They will be the ones that invest in enabling their people, through training, experimentation, and a culture that encourages learning. Ultimately, AI transformation is not about replacing human expertise. It is about empowering people to work smarter, focus on higher-value tasks, and rethink how work gets done.
About Niv Liran
Niv Liran is the Chief Product and Technology Officer at Unzer, where he leads the company’s product and engineering teams and shapes its technology strategy. Under his leadership, Unzer has built a single platform that brings together online and in-store sales, giving businesses a complete, real-time view of their customers and operations. He also guides Unzer’s work in artificial intelligence, overseeing the company’s AI roadmap and initiatives.
