IT leaders gained key insights from industry experts at Canvas 25 regarding AI transformation. A primary theme was the importance of focusing on customer needs rather than just the allure of AI technology. GitHub's Andrew Bauer emphasized the risk of jumping on the AI bandwagon without understanding user requirements. He stated that simply labeling something as "Powered by AI" doesn't guarantee value. Instead, companies should start by listening to their customers to align AI solutions with real problems. AWS's Eduardo Ordax echoed this, warning against an AI-first mentality and advocating for a customer-first strategy. Culture change emerged as another critical factor. Des Traynor from Intercom shared how his company faced a dramatic shift after realizing the impact of AI tools like ChatGPT. The challenge wasn't technical but cultural. Traynor highlighted the need for a determined approach to change, stressing that leaders should set high expectations for AI integration and be willing to discard outdated processes. Red Hat's Jan Mark Holzer reinforced this, pointing out that technology is the easier part; getting people onboard is where the difficulty lies. Data quality is foundational for successful AI implementation. Holzer pointed out that poor data can lead to misleading outputs from AI systems, a sentiment supported by Ordax’s warning that “garbage in, garbage out” holds true for AI. Industries with a solid data strategy, such as finance and healthcare, are better equipped for AI success. Holzer recommended creating a centralized data layer with strict governance and quality standards to avoid pitfalls. The overall takeaway is that IT leaders need to act as strategic partners to their CEOs, focusing on delivering measurable business value rather than just new tools.