After realizing the high costs of hiring external AI experts, some CIOs have devised methods to cultivate AI skills internally—not only within IT but throughout the entire organization. Early adopters have identified four distinct approaches that any company should consider for AI training programs. **Increasing Office Productivity** Arco, a construction services company, initiated AI training focused on using Microsoft Copilot to transcribe meetings, generate action items, and integrate them into Microsoft Planner. The first sessions targeted a small group of executives and assistants, with Microsoft experts leading the training. Ensuring privacy was a priority, so meeting transcripts stayed within the company, according to Robin Patra, Arco’s director of data and AI. Success was measured by three metrics: the frequency of Copilot activation compared to the number of meetings attended, the integration of Copilot output into workflows, and attendee satisfaction surveys. After a successful pilot in October 2024, Arco expanded the tool and made training mandatory for all 4, 000 employees by November. The course, called AI 101, is now an online, five-hour program covering AI fundamentals. **Improving Core Functions** Building on this, Arco launched a second course, AI 102, a voluntary five-day online program focusing on aligning AI with business challenges across construction processes—from estimation and design to project management and execution. About two-thirds of the workforce has completed this course, which requires participants to submit at least one innovation idea to the company’s portal. The innovation and engineering teams review proposals, sometimes engaging with contributors for further exploration. For example, a legal team member suggested using large language models (LLMs) to expedite case review by identifying similar past cases, enabling quicker responses. This led to implementing a legal AI tool to support document analysis and response drafting. **Enthusiast Training with Low-Code/No-Code Tools** Arco also created a third program targeting AI enthusiasts interested in building applications using low-code and no-code platforms tailored to construction services. This hands-on training, held quarterly at the St. Louis headquarters and led by external instructors, has involved around 80 participants eager to develop AI prompts and solutions. **Developing AI Skills Across the Organization** Engineering companies often lead in tech adoption. Lexmark exemplifies this by partnering with North Carolina State University four years ago to establish the AI Academy, allowing employees to pursue tuition-free degrees, as explained by Vishal Gupta, CITO.

Initially having just five AI data scientists, Lexmark now has 100 who completed four key training tracks. Training extends beyond specialists to employees in HR, finance, manufacturing, and other sectors. Even those without programming experience learn Python to build AI applications. Volunteers attend three-hour classes four evenings weekly over a year, paired with mentors and assigned projects aligned with company goals. Gupta notes that nobody has dropped out, and attrition is minimal, as employees appreciate the opportunity to develop skills and apply them. So far, six cohorts have graduated, equipping Lexmark not only with skilled personnel but also with practical AI use cases sourced from diverse business areas. Graduates can identify AI-solvable problems across manufacturing, customer service, sales, and more. **Building an AI Culture** Marc Booker, vice provost of strategy at the University of Phoenix, advocates learning AI through hands-on experience in communities of practice and mentorships. These forums bring technologists and less experienced colleagues together to share ideas and collaborate on real problems. Communities typically focus on skill development in areas like machine learning and LLMs and encourage diverse team compositions. Such communities facilitate change management by building skills and reducing fear of AI adoption. Leadership involvement and designated facilitators are essential, while mentorship often emerges naturally. Booker observes that business professionals sometimes evolve into technologists through participation. Lexmark also emphasizes culture building beyond technical training. Last year, it introduced AI Foundations, a course designed to reduce apprehension and encourage early adoption. They anticipated around 1, 000 participants but 5, 000 of their 7, 000 employees enrolled within two months, demonstrating widespread enthusiasm. Forward-thinking CIOs and IT leaders view this enthusiasm as an opportunity to not only deploy new AI tools but also to nurture innovation and empower employees to solve problems creatively. By focusing training programs on clear outcomes, IT heads can unlock AI’s full potential and prepare a workforce equipped for future challenges. **More on building AI skills:**