The rapid ascent of artificial intelligence has left many professionals grappling with uncertainty. Concerns range from shifting job roles and the potential elimination of certain positions to the fear of being outpaced by competitors who adopt AI more aggressively. When financial resources for AI experimentation are scarce, those anxieties only deepen.
The good news, according to technology leaders across industries, is that meaningful AI exploration doesn’t require a massive budget. Here are five practical strategies for dipping into AI cost-effectively.
1. Leverage What You Already Have
Nick Pearson, CIO at Ricoh Europe, emphasizes the importance of avoiding unnecessary reinvention. Many organizations already possess tools that can support AI explorations without additional investment.
“This goes back to where I am right now, which is utilizing and leveraging what there already is—and that approach is actually getting easier to a certain degree,” Pearson explains.
He points to Microsoft 365 as a prime example. Most organizations running Microsoft environments already have Copilot embedded within their existing licensing. For professionals already inside the workplace, there are capabilities ready to be activated.
Pearson notes that affordability varies by organization, and the key is understanding what’s natural and sustainable for your specific business context.
2. Tap Into Open-Source Capabilities
Joel Hron, CTO at Thomson Reuters, advises professionals to look beyond commercial solutions. His organization combines in-house models with off-the-shelf tools, but he emphasizes that open-source alternatives offer a viable path for those with constrained budgets.
“I honestly don’t think I would focus on anything different” if resources were limited, Hron says. “I wouldn’t go training my own models and things like that if I didn’t have resources to do it.”
The open-source AI community, he notes, is as prolific as the commercial IT industry. Many projects can be explored on virtually no budget, helping professionals build intuition around AI’s capabilities and trajectory.
“That intuition, I think, is valuable for anybody right now,” Hron adds.
3. Exploit Cloud Flexibility
Huy Dao, director of data and machine learning platform at Booking.com, advocates for cloud-based approaches as the natural fit for cost-conscious AI exploration.
“With the cloud, you don’t have to invest so much money upfront,” Dao explains. “If your business idea becomes successful, you pay more. If your business idea isn’t growing as quickly, then you don’t pay as much.”
Booking.com uses Snowflake as its cloud data platform, with costs scaling directly with usage. Dao notes that this model fundamentally changes the economics of innovation.
“In the past, you had to be a bigger company to get involved,” he says. “Nowadays, you can start with a very small, cloud-based or OpenAI subscription, and you can get started there.”
His advice: leverage cloud services to maintain flexibility, and move past any lingering skepticism about AI adoption.
4. Focus on Desired Outcomes
Musidora Jorgensen, UK & Ireland country leader at Freshworks, cautions against pursuing AI for its own sake.
“AI for the sake of it doesn’t drive the outcomes that people want,” Jorgensen says. “So, home in on the problem you’re trying to solve, the outcome you’re looking for, and the efficiencies that AI can bring.”
Once the right tools are selected, she emphasizes the importance of supporting teams as they integrate AI into daily workflows. When employees are empowered to adopt AI effectively, efficiency gains follow naturally, freeing up capacity for higher-value strategic work.
5. Stay Flexible to Change
Thierry Martin, head of enterprise data and analytics at Toyota Motor Europe, suggests that smaller organizations and those with lean IT estates actually possess a natural advantage: agility.
“Smaller companies can have an advantage, because they don’t have the anchor of the legacy systems that bigger, older firms are pulling,” Martin observes.
His advice for cost-conscious AI explorations is to embrace imperfection. “Don’t target 100%, target 80%. Don’t shoot for the stars, because the moon is moving.”
Martin cites the rapid emergence of Anthropic’s Model Context Protocol (MCP) as an example of why flexibility matters. An open standard that connects AI applications to external systems, MCP gained traction quickly—and organizations that had committed to rigid, long-term plans found themselves needing to pivot.
“The goal is always moving further away, so I think 80% is always good,” Martin says. “As you are moving, some new targets will arrive, so try to keep the road to your goal as straight as possible, so that you’re not confusing people.”
