Enterprises keep betting on coding agents despite lackluster results

Enterprises are rapidly increasing spending on coding agents and agentic software development, but the business payoff is still uneven: only about a quarter of companies report meaningful delivery acceleration, while 30% say productivity declined after adoption. For CIOs, the strategic takeaway is that agentic AI is becoming a core operating model issue—not just a tool choice—requiring tighter governance, better documentation and context systems, and a shift toward smaller teams that supervise AI execution without sacrificing quality, maintainability, or control.