Drew Cohen Tells Us Whether Software Stocks Are Dead (Agents, Models, Moats, and More)

The software industry faces significant narrative pressure regarding AI-driven disruption, yet the threat of enterprises replacing established systems with in-house "vibe coded" tools remains largely overstated. While AI lowers development costs, the complexity of governance, security, and maintenance ensures that core enterprise software remains integral. The more significant structural risk involves the "outer harness"—the user interface where AI work originates—which could disintermediate traditional software platforms if AI models become the primary dashboard for business operations. Consumer software and niche vertical SaaS applications are particularly vulnerable, whereas platforms with proprietary data or established trust layers, such as Amazon or Airbnb, maintain greater resilience. Meanwhile, ad-tech players like Meta benefit from improved ad matching and creative generation, though search-dependent businesses face long-term pressure as AI agents shift user habits away from traditional search queries.