Vibe coding: here to stay or just the latest enterprise fad? (www.techradar.com)

Gartner’s new report and a flurry of high-profile demos have pushed “vibe coding” — agentic, GenAI-driven rapid app creation — into the spotlight, with analysts estimating up to 40% of new enterprise production software could use these techniques by 2028. Proponents point to clear benefits: dramatic speed, lower barriers for non-developers (a father-daughter duo built a business site in six hours), and startup wins—Swedish AI company Lovable claims minute-scale product creation and projects $1B ARR—while small-business users report big productivity and income gains. But the trend exposes real technical and governance gaps for enterprise use. Large language models excel at pattern matching and plausible generation, not intent or context, which can lead to catastrophic actions (Replit’s AI reportedly deleted a client database). Vibe coding struggles with non-functional requirements—security, compliance, auditability, scalability—and lacks built-in guardrails. For enterprises to safely adopt these tools they must update SDLCs, embed data engineers and human-in-the-loop checks, enforce governance and testing, and treat vibe coding as an accelerant rather than a replacement for skilled engineering. Absent those controls, the speed gains risk short-term churn but long-term operational liability.